forked from p30928647/excelize
17957 lines
559 KiB
Go
17957 lines
559 KiB
Go
// Copyright 2016 - 2023 The excelize Authors. All rights reserved. Use of
|
|
// this source code is governed by a BSD-style license that can be found in
|
|
// the LICENSE file.
|
|
//
|
|
// Package excelize providing a set of functions that allow you to write to and
|
|
// read from XLAM / XLSM / XLSX / XLTM / XLTX files. Supports reading and
|
|
// writing spreadsheet documents generated by Microsoft Excel™ 2007 and later.
|
|
// Supports complex components by high compatibility, and provided streaming
|
|
// API for generating or reading data from a worksheet with huge amounts of
|
|
// data. This library needs Go version 1.16 or later.
|
|
|
|
package excelize
|
|
|
|
import (
|
|
"bytes"
|
|
"container/list"
|
|
"errors"
|
|
"fmt"
|
|
"math"
|
|
"math/big"
|
|
"math/cmplx"
|
|
"math/rand"
|
|
"net/url"
|
|
"reflect"
|
|
"regexp"
|
|
"sort"
|
|
"strconv"
|
|
"strings"
|
|
"sync"
|
|
"time"
|
|
"unicode"
|
|
"unicode/utf8"
|
|
"unsafe"
|
|
|
|
"github.com/xuri/efp"
|
|
"golang.org/x/text/language"
|
|
"golang.org/x/text/message"
|
|
)
|
|
|
|
const (
|
|
// Excel formula errors
|
|
formulaErrorDIV = "#DIV/0!"
|
|
formulaErrorNAME = "#NAME?"
|
|
formulaErrorNA = "#N/A"
|
|
formulaErrorNUM = "#NUM!"
|
|
formulaErrorVALUE = "#VALUE!"
|
|
formulaErrorREF = "#REF!"
|
|
formulaErrorNULL = "#NULL!"
|
|
formulaErrorSPILL = "#SPILL!"
|
|
formulaErrorCALC = "#CALC!"
|
|
formulaErrorGETTINGDATA = "#GETTING_DATA"
|
|
// Formula criteria condition enumeration
|
|
_ byte = iota
|
|
criteriaEq
|
|
criteriaLe
|
|
criteriaGe
|
|
criteriaNe
|
|
criteriaL
|
|
criteriaG
|
|
criteriaErr
|
|
criteriaRegexp
|
|
|
|
categoryWeightAndMass
|
|
categoryDistance
|
|
categoryTime
|
|
categoryPressure
|
|
categoryForce
|
|
categoryEnergy
|
|
categoryPower
|
|
categoryMagnetism
|
|
categoryTemperature
|
|
categoryVolumeAndLiquidMeasure
|
|
categoryArea
|
|
categoryInformation
|
|
categorySpeed
|
|
|
|
matchModeExact = 0
|
|
matchModeMinGreater = 1
|
|
matchModeMaxLess = -1
|
|
matchModeWildcard = 2
|
|
|
|
searchModeLinear = 1
|
|
searchModeReverseLinear = -1
|
|
searchModeAscBinary = 2
|
|
searchModeDescBinary = -2
|
|
|
|
maxFinancialIterations = 128
|
|
financialPrecision = 1.0e-08
|
|
// Date and time format regular expressions
|
|
monthRe = `((jan|january)|(feb|february)|(mar|march)|(apr|april)|(may)|(jun|june)|(jul|july)|(aug|august)|(sep|september)|(oct|october)|(nov|november)|(dec|december))`
|
|
df1 = `(([0-9])+)/(([0-9])+)/(([0-9])+)`
|
|
df2 = monthRe + ` (([0-9])+), (([0-9])+)`
|
|
df3 = `(([0-9])+)-(([0-9])+)-(([0-9])+)`
|
|
df4 = `(([0-9])+)-` + monthRe + `-(([0-9])+)`
|
|
datePrefix = `^((` + df1 + `|` + df2 + `|` + df3 + `|` + df4 + `) )?`
|
|
tfhh = `(([0-9])+) (am|pm)`
|
|
tfhhmm = `(([0-9])+):(([0-9])+)( (am|pm))?`
|
|
tfmmss = `(([0-9])+):(([0-9])+\.([0-9])+)( (am|pm))?`
|
|
tfhhmmss = `(([0-9])+):(([0-9])+):(([0-9])+(\.([0-9])+)?)( (am|pm))?`
|
|
timeSuffix = `( (` + tfhh + `|` + tfhhmm + `|` + tfmmss + `|` + tfhhmmss + `))?$`
|
|
)
|
|
|
|
var (
|
|
// tokenPriority defined basic arithmetic operator priority
|
|
tokenPriority = map[string]int{
|
|
"^": 5,
|
|
"*": 4,
|
|
"/": 4,
|
|
"+": 3,
|
|
"-": 3,
|
|
"=": 2,
|
|
"<>": 2,
|
|
"<": 2,
|
|
"<=": 2,
|
|
">": 2,
|
|
">=": 2,
|
|
"&": 1,
|
|
}
|
|
month2num = map[string]int{
|
|
"january": 1,
|
|
"february": 2,
|
|
"march": 3,
|
|
"april": 4,
|
|
"may": 5,
|
|
"june": 6,
|
|
"july": 7,
|
|
"august": 8,
|
|
"september": 9,
|
|
"october": 10,
|
|
"november": 11,
|
|
"december": 12,
|
|
"jan": 1,
|
|
"feb": 2,
|
|
"mar": 3,
|
|
"apr": 4,
|
|
"jun": 6,
|
|
"jul": 7,
|
|
"aug": 8,
|
|
"sep": 9,
|
|
"oct": 10,
|
|
"nov": 11,
|
|
"dec": 12,
|
|
}
|
|
dateFormats = map[string]*regexp.Regexp{
|
|
"mm/dd/yy": regexp.MustCompile(`^` + df1 + timeSuffix),
|
|
"mm dd, yy": regexp.MustCompile(`^` + df2 + timeSuffix),
|
|
"yy-mm-dd": regexp.MustCompile(`^` + df3 + timeSuffix),
|
|
"yy-mmStr-dd": regexp.MustCompile(`^` + df4 + timeSuffix),
|
|
}
|
|
timeFormats = map[string]*regexp.Regexp{
|
|
"hh": regexp.MustCompile(datePrefix + tfhh + `$`),
|
|
"hh:mm": regexp.MustCompile(datePrefix + tfhhmm + `$`),
|
|
"mm:ss": regexp.MustCompile(datePrefix + tfmmss + `$`),
|
|
"hh:mm:ss": regexp.MustCompile(datePrefix + tfhhmmss + `$`),
|
|
}
|
|
dateOnlyFormats = []*regexp.Regexp{
|
|
regexp.MustCompile(`^` + df1 + `$`),
|
|
regexp.MustCompile(`^` + df2 + `$`),
|
|
regexp.MustCompile(`^` + df3 + `$`),
|
|
regexp.MustCompile(`^` + df4 + `$`),
|
|
}
|
|
addressFmtMaps = map[string]func(col, row int) (string, error){
|
|
"1_TRUE": func(col, row int) (string, error) {
|
|
return CoordinatesToCellName(col, row, true)
|
|
},
|
|
"1_FALSE": func(col, row int) (string, error) {
|
|
return fmt.Sprintf("R%dC%d", row, col), nil
|
|
},
|
|
"2_TRUE": func(col, row int) (string, error) {
|
|
column, err := ColumnNumberToName(col)
|
|
if err != nil {
|
|
return "", err
|
|
}
|
|
return fmt.Sprintf("%s$%d", column, row), nil
|
|
},
|
|
"2_FALSE": func(col, row int) (string, error) {
|
|
return fmt.Sprintf("R%dC[%d]", row, col), nil
|
|
},
|
|
"3_TRUE": func(col, row int) (string, error) {
|
|
column, err := ColumnNumberToName(col)
|
|
if err != nil {
|
|
return "", err
|
|
}
|
|
return fmt.Sprintf("$%s%d", column, row), nil
|
|
},
|
|
"3_FALSE": func(col, row int) (string, error) {
|
|
return fmt.Sprintf("R[%d]C%d", row, col), nil
|
|
},
|
|
"4_TRUE": func(col, row int) (string, error) {
|
|
return CoordinatesToCellName(col, row, false)
|
|
},
|
|
"4_FALSE": func(col, row int) (string, error) {
|
|
return fmt.Sprintf("R[%d]C[%d]", row, col), nil
|
|
},
|
|
}
|
|
)
|
|
|
|
// calcContext defines the formula execution context.
|
|
type calcContext struct {
|
|
sync.Mutex
|
|
entry string
|
|
maxCalcIterations uint
|
|
iterations map[string]uint
|
|
iterationsCache map[string]formulaArg
|
|
}
|
|
|
|
// cellRef defines the structure of a cell reference.
|
|
type cellRef struct {
|
|
Col int
|
|
Row int
|
|
Sheet string
|
|
}
|
|
|
|
// cellRef defines the structure of a cell range.
|
|
type cellRange struct {
|
|
From cellRef
|
|
To cellRef
|
|
}
|
|
|
|
// formulaCriteria defined formula criteria parser result.
|
|
type formulaCriteria struct {
|
|
Type byte
|
|
Condition string
|
|
}
|
|
|
|
// ArgType is the type of formula argument type.
|
|
type ArgType byte
|
|
|
|
// Formula argument types enumeration.
|
|
const (
|
|
ArgUnknown ArgType = iota
|
|
ArgNumber
|
|
ArgString
|
|
ArgList
|
|
ArgMatrix
|
|
ArgError
|
|
ArgEmpty
|
|
)
|
|
|
|
// formulaArg is the argument of a formula or function.
|
|
type formulaArg struct {
|
|
SheetName string
|
|
Number float64
|
|
String string
|
|
List []formulaArg
|
|
Matrix [][]formulaArg
|
|
Boolean bool
|
|
Error string
|
|
Type ArgType
|
|
cellRefs, cellRanges *list.List
|
|
}
|
|
|
|
// Value returns a string data type of the formula argument.
|
|
func (fa formulaArg) Value() (value string) {
|
|
switch fa.Type {
|
|
case ArgNumber:
|
|
if fa.Boolean {
|
|
if fa.Number == 0 {
|
|
return "FALSE"
|
|
}
|
|
return "TRUE"
|
|
}
|
|
return fmt.Sprintf("%g", fa.Number)
|
|
case ArgString:
|
|
return fa.String
|
|
case ArgError:
|
|
return fa.Error
|
|
}
|
|
return
|
|
}
|
|
|
|
// ToNumber returns a formula argument with number data type.
|
|
func (fa formulaArg) ToNumber() formulaArg {
|
|
var n float64
|
|
var err error
|
|
switch fa.Type {
|
|
case ArgString:
|
|
n, err = strconv.ParseFloat(fa.String, 64)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
case ArgNumber:
|
|
n = fa.Number
|
|
}
|
|
return newNumberFormulaArg(n)
|
|
}
|
|
|
|
// ToBool returns a formula argument with boolean data type.
|
|
func (fa formulaArg) ToBool() formulaArg {
|
|
var b bool
|
|
var err error
|
|
switch fa.Type {
|
|
case ArgString:
|
|
b, err = strconv.ParseBool(fa.String)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
case ArgNumber:
|
|
if fa.Boolean && fa.Number == 1 {
|
|
b = true
|
|
}
|
|
}
|
|
return newBoolFormulaArg(b)
|
|
}
|
|
|
|
// ToList returns a formula argument with array data type.
|
|
func (fa formulaArg) ToList() []formulaArg {
|
|
switch fa.Type {
|
|
case ArgMatrix:
|
|
var args []formulaArg
|
|
for _, row := range fa.Matrix {
|
|
args = append(args, row...)
|
|
}
|
|
return args
|
|
case ArgList:
|
|
return fa.List
|
|
case ArgNumber, ArgString, ArgError, ArgUnknown:
|
|
return []formulaArg{fa}
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// formulaFuncs is the type of the formula functions.
|
|
type formulaFuncs struct {
|
|
f *File
|
|
ctx *calcContext
|
|
sheet, cell string
|
|
}
|
|
|
|
// CalcCellValue provides a function to get calculated cell value. This feature
|
|
// is currently in working processing. Iterative calculation, implicit
|
|
// intersection, explicit intersection, array formula, table formula and some
|
|
// other formulas are not supported currently.
|
|
//
|
|
// Supported formula functions:
|
|
//
|
|
// ABS
|
|
// ACCRINT
|
|
// ACCRINTM
|
|
// ACOS
|
|
// ACOSH
|
|
// ACOT
|
|
// ACOTH
|
|
// ADDRESS
|
|
// AGGREGATE
|
|
// AMORDEGRC
|
|
// AMORLINC
|
|
// AND
|
|
// ARABIC
|
|
// ASIN
|
|
// ASINH
|
|
// ATAN
|
|
// ATAN2
|
|
// ATANH
|
|
// AVEDEV
|
|
// AVERAGE
|
|
// AVERAGEA
|
|
// AVERAGEIF
|
|
// AVERAGEIFS
|
|
// BASE
|
|
// BESSELI
|
|
// BESSELJ
|
|
// BESSELK
|
|
// BESSELY
|
|
// BETADIST
|
|
// BETA.DIST
|
|
// BETAINV
|
|
// BETA.INV
|
|
// BIN2DEC
|
|
// BIN2HEX
|
|
// BIN2OCT
|
|
// BINOMDIST
|
|
// BINOM.DIST
|
|
// BINOM.DIST.RANGE
|
|
// BINOM.INV
|
|
// BITAND
|
|
// BITLSHIFT
|
|
// BITOR
|
|
// BITRSHIFT
|
|
// BITXOR
|
|
// CEILING
|
|
// CEILING.MATH
|
|
// CEILING.PRECISE
|
|
// CHAR
|
|
// CHIDIST
|
|
// CHIINV
|
|
// CHITEST
|
|
// CHISQ.DIST
|
|
// CHISQ.DIST.RT
|
|
// CHISQ.INV
|
|
// CHISQ.INV.RT
|
|
// CHISQ.TEST
|
|
// CHOOSE
|
|
// CLEAN
|
|
// CODE
|
|
// COLUMN
|
|
// COLUMNS
|
|
// COMBIN
|
|
// COMBINA
|
|
// COMPLEX
|
|
// CONCAT
|
|
// CONCATENATE
|
|
// CONFIDENCE
|
|
// CONFIDENCE.NORM
|
|
// CONFIDENCE.T
|
|
// CONVERT
|
|
// CORREL
|
|
// COS
|
|
// COSH
|
|
// COT
|
|
// COTH
|
|
// COUNT
|
|
// COUNTA
|
|
// COUNTBLANK
|
|
// COUNTIF
|
|
// COUNTIFS
|
|
// COUPDAYBS
|
|
// COUPDAYS
|
|
// COUPDAYSNC
|
|
// COUPNCD
|
|
// COUPNUM
|
|
// COUPPCD
|
|
// COVAR
|
|
// COVARIANCE.P
|
|
// COVARIANCE.S
|
|
// CRITBINOM
|
|
// CSC
|
|
// CSCH
|
|
// CUMIPMT
|
|
// CUMPRINC
|
|
// DATE
|
|
// DATEDIF
|
|
// DATEVALUE
|
|
// DAVERAGE
|
|
// DAY
|
|
// DAYS
|
|
// DAYS360
|
|
// DB
|
|
// DCOUNT
|
|
// DCOUNTA
|
|
// DDB
|
|
// DEC2BIN
|
|
// DEC2HEX
|
|
// DEC2OCT
|
|
// DECIMAL
|
|
// DEGREES
|
|
// DELTA
|
|
// DEVSQ
|
|
// DGET
|
|
// DISC
|
|
// DMAX
|
|
// DMIN
|
|
// DOLLARDE
|
|
// DOLLARFR
|
|
// DPRODUCT
|
|
// DSTDEV
|
|
// DSTDEVP
|
|
// DSUM
|
|
// DURATION
|
|
// DVAR
|
|
// DVARP
|
|
// EFFECT
|
|
// EDATE
|
|
// ENCODEURL
|
|
// EOMONTH
|
|
// ERF
|
|
// ERF.PRECISE
|
|
// ERFC
|
|
// ERFC.PRECISE
|
|
// ERROR.TYPE
|
|
// EUROCONVERT
|
|
// EVEN
|
|
// EXACT
|
|
// EXP
|
|
// EXPON.DIST
|
|
// EXPONDIST
|
|
// FACT
|
|
// FACTDOUBLE
|
|
// FALSE
|
|
// F.DIST
|
|
// F.DIST.RT
|
|
// FDIST
|
|
// FIND
|
|
// FINDB
|
|
// F.INV
|
|
// F.INV.RT
|
|
// FINV
|
|
// FISHER
|
|
// FISHERINV
|
|
// FIXED
|
|
// FLOOR
|
|
// FLOOR.MATH
|
|
// FLOOR.PRECISE
|
|
// FORMULATEXT
|
|
// F.TEST
|
|
// FTEST
|
|
// FV
|
|
// FVSCHEDULE
|
|
// GAMMA
|
|
// GAMMA.DIST
|
|
// GAMMADIST
|
|
// GAMMA.INV
|
|
// GAMMAINV
|
|
// GAMMALN
|
|
// GAMMALN.PRECISE
|
|
// GAUSS
|
|
// GCD
|
|
// GEOMEAN
|
|
// GESTEP
|
|
// GROWTH
|
|
// HARMEAN
|
|
// HEX2BIN
|
|
// HEX2DEC
|
|
// HEX2OCT
|
|
// HLOOKUP
|
|
// HOUR
|
|
// HYPERLINK
|
|
// HYPGEOM.DIST
|
|
// HYPGEOMDIST
|
|
// IF
|
|
// IFERROR
|
|
// IFNA
|
|
// IFS
|
|
// IMABS
|
|
// IMAGINARY
|
|
// IMARGUMENT
|
|
// IMCONJUGATE
|
|
// IMCOS
|
|
// IMCOSH
|
|
// IMCOT
|
|
// IMCSC
|
|
// IMCSCH
|
|
// IMDIV
|
|
// IMEXP
|
|
// IMLN
|
|
// IMLOG10
|
|
// IMLOG2
|
|
// IMPOWER
|
|
// IMPRODUCT
|
|
// IMREAL
|
|
// IMSEC
|
|
// IMSECH
|
|
// IMSIN
|
|
// IMSINH
|
|
// IMSQRT
|
|
// IMSUB
|
|
// IMSUM
|
|
// IMTAN
|
|
// INDEX
|
|
// INDIRECT
|
|
// INT
|
|
// INTRATE
|
|
// IPMT
|
|
// IRR
|
|
// ISBLANK
|
|
// ISERR
|
|
// ISERROR
|
|
// ISEVEN
|
|
// ISFORMULA
|
|
// ISLOGICAL
|
|
// ISNA
|
|
// ISNONTEXT
|
|
// ISNUMBER
|
|
// ISODD
|
|
// ISREF
|
|
// ISTEXT
|
|
// ISO.CEILING
|
|
// ISOWEEKNUM
|
|
// ISPMT
|
|
// KURT
|
|
// LARGE
|
|
// LCM
|
|
// LEFT
|
|
// LEFTB
|
|
// LEN
|
|
// LENB
|
|
// LN
|
|
// LOG
|
|
// LOG10
|
|
// LOGINV
|
|
// LOGNORM.DIST
|
|
// LOGNORMDIST
|
|
// LOGNORM.INV
|
|
// LOOKUP
|
|
// LOWER
|
|
// MATCH
|
|
// MAX
|
|
// MAXA
|
|
// MAXIFS
|
|
// MDETERM
|
|
// MDURATION
|
|
// MEDIAN
|
|
// MID
|
|
// MIDB
|
|
// MIN
|
|
// MINA
|
|
// MINIFS
|
|
// MINUTE
|
|
// MINVERSE
|
|
// MIRR
|
|
// MMULT
|
|
// MOD
|
|
// MODE
|
|
// MODE.MULT
|
|
// MODE.SNGL
|
|
// MONTH
|
|
// MROUND
|
|
// MULTINOMIAL
|
|
// MUNIT
|
|
// N
|
|
// NA
|
|
// NEGBINOM.DIST
|
|
// NEGBINOMDIST
|
|
// NETWORKDAYS
|
|
// NETWORKDAYS.INTL
|
|
// NOMINAL
|
|
// NORM.DIST
|
|
// NORMDIST
|
|
// NORM.INV
|
|
// NORMINV
|
|
// NORM.S.DIST
|
|
// NORMSDIST
|
|
// NORM.S.INV
|
|
// NORMSINV
|
|
// NOT
|
|
// NOW
|
|
// NPER
|
|
// NPV
|
|
// OCT2BIN
|
|
// OCT2DEC
|
|
// OCT2HEX
|
|
// ODD
|
|
// ODDFPRICE
|
|
// OR
|
|
// PDURATION
|
|
// PEARSON
|
|
// PERCENTILE.EXC
|
|
// PERCENTILE.INC
|
|
// PERCENTILE
|
|
// PERCENTRANK.EXC
|
|
// PERCENTRANK.INC
|
|
// PERCENTRANK
|
|
// PERMUT
|
|
// PERMUTATIONA
|
|
// PHI
|
|
// PI
|
|
// PMT
|
|
// POISSON.DIST
|
|
// POISSON
|
|
// POWER
|
|
// PPMT
|
|
// PRICE
|
|
// PRICEDISC
|
|
// PRICEMAT
|
|
// PRODUCT
|
|
// PROPER
|
|
// PV
|
|
// QUARTILE
|
|
// QUARTILE.EXC
|
|
// QUARTILE.INC
|
|
// QUOTIENT
|
|
// RADIANS
|
|
// RAND
|
|
// RANDBETWEEN
|
|
// RANK
|
|
// RANK.EQ
|
|
// RATE
|
|
// RECEIVED
|
|
// REPLACE
|
|
// REPLACEB
|
|
// REPT
|
|
// RIGHT
|
|
// RIGHTB
|
|
// ROMAN
|
|
// ROUND
|
|
// ROUNDDOWN
|
|
// ROUNDUP
|
|
// ROW
|
|
// ROWS
|
|
// RRI
|
|
// RSQ
|
|
// SEC
|
|
// SECH
|
|
// SECOND
|
|
// SERIESSUM
|
|
// SHEET
|
|
// SHEETS
|
|
// SIGN
|
|
// SIN
|
|
// SINH
|
|
// SKEW
|
|
// SKEW.P
|
|
// SLN
|
|
// SLOPE
|
|
// SMALL
|
|
// SQRT
|
|
// SQRTPI
|
|
// STANDARDIZE
|
|
// STDEV
|
|
// STDEV.P
|
|
// STDEV.S
|
|
// STDEVA
|
|
// STDEVP
|
|
// STDEVPA
|
|
// STEYX
|
|
// SUBSTITUTE
|
|
// SUBTOTAL
|
|
// SUM
|
|
// SUMIF
|
|
// SUMIFS
|
|
// SUMPRODUCT
|
|
// SUMSQ
|
|
// SUMX2MY2
|
|
// SUMX2PY2
|
|
// SUMXMY2
|
|
// SWITCH
|
|
// SYD
|
|
// T
|
|
// TAN
|
|
// TANH
|
|
// TBILLEQ
|
|
// TBILLPRICE
|
|
// TBILLYIELD
|
|
// T.DIST
|
|
// T.DIST.2T
|
|
// T.DIST.RT
|
|
// TDIST
|
|
// TEXTJOIN
|
|
// TIME
|
|
// TIMEVALUE
|
|
// T.INV
|
|
// T.INV.2T
|
|
// TINV
|
|
// TODAY
|
|
// TRANSPOSE
|
|
// TREND
|
|
// TRIM
|
|
// TRIMMEAN
|
|
// TRUE
|
|
// TRUNC
|
|
// T.TEST
|
|
// TTEST
|
|
// TYPE
|
|
// UNICHAR
|
|
// UNICODE
|
|
// UPPER
|
|
// VALUE
|
|
// VAR
|
|
// VAR.P
|
|
// VAR.S
|
|
// VARA
|
|
// VARP
|
|
// VARPA
|
|
// VDB
|
|
// VLOOKUP
|
|
// WEEKDAY
|
|
// WEEKNUM
|
|
// WEIBULL
|
|
// WEIBULL.DIST
|
|
// WORKDAY
|
|
// WORKDAY.INTL
|
|
// XIRR
|
|
// XLOOKUP
|
|
// XNPV
|
|
// XOR
|
|
// YEAR
|
|
// YEARFRAC
|
|
// YIELD
|
|
// YIELDDISC
|
|
// YIELDMAT
|
|
// Z.TEST
|
|
// ZTEST
|
|
func (f *File) CalcCellValue(sheet, cell string, opts ...Options) (result string, err error) {
|
|
var (
|
|
rawCellValue = getOptions(opts...).RawCellValue
|
|
styleIdx int
|
|
token formulaArg
|
|
)
|
|
if token, err = f.calcCellValue(&calcContext{
|
|
entry: fmt.Sprintf("%s!%s", sheet, cell),
|
|
maxCalcIterations: getOptions(opts...).MaxCalcIterations,
|
|
iterations: make(map[string]uint),
|
|
iterationsCache: make(map[string]formulaArg),
|
|
}, sheet, cell); err != nil {
|
|
return
|
|
}
|
|
if !rawCellValue {
|
|
styleIdx, _ = f.GetCellStyle(sheet, cell)
|
|
}
|
|
result = token.Value()
|
|
if isNum, precision, decimal := isNumeric(result); isNum {
|
|
if precision > 15 {
|
|
result, err = f.formattedValue(styleIdx, strings.ToUpper(strconv.FormatFloat(decimal, 'G', 15, 64)), rawCellValue)
|
|
return
|
|
}
|
|
if !strings.HasPrefix(result, "0") {
|
|
result, err = f.formattedValue(styleIdx, strings.ToUpper(strconv.FormatFloat(decimal, 'f', -1, 64)), rawCellValue)
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// calcCellValue calculate cell value by given context, worksheet name and cell
|
|
// reference.
|
|
func (f *File) calcCellValue(ctx *calcContext, sheet, cell string) (result formulaArg, err error) {
|
|
var formula string
|
|
if formula, err = f.GetCellFormula(sheet, cell); err != nil {
|
|
return
|
|
}
|
|
ps := efp.ExcelParser()
|
|
tokens := ps.Parse(formula)
|
|
if tokens == nil {
|
|
return
|
|
}
|
|
result, err = f.evalInfixExp(ctx, sheet, cell, tokens)
|
|
return
|
|
}
|
|
|
|
// getPriority calculate arithmetic operator priority.
|
|
func getPriority(token efp.Token) (pri int) {
|
|
pri = tokenPriority[token.TValue]
|
|
if token.TValue == "-" && token.TType == efp.TokenTypeOperatorPrefix {
|
|
pri = 6
|
|
}
|
|
if isBeginParenthesesToken(token) { // (
|
|
pri = 0
|
|
}
|
|
return
|
|
}
|
|
|
|
// newNumberFormulaArg constructs a number formula argument.
|
|
func newNumberFormulaArg(n float64) formulaArg {
|
|
if math.IsNaN(n) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return formulaArg{Type: ArgNumber, Number: n}
|
|
}
|
|
|
|
// newStringFormulaArg constructs a string formula argument.
|
|
func newStringFormulaArg(s string) formulaArg {
|
|
return formulaArg{Type: ArgString, String: s}
|
|
}
|
|
|
|
// newMatrixFormulaArg constructs a matrix formula argument.
|
|
func newMatrixFormulaArg(m [][]formulaArg) formulaArg {
|
|
return formulaArg{Type: ArgMatrix, Matrix: m}
|
|
}
|
|
|
|
// newListFormulaArg create a list formula argument.
|
|
func newListFormulaArg(l []formulaArg) formulaArg {
|
|
return formulaArg{Type: ArgList, List: l}
|
|
}
|
|
|
|
// newBoolFormulaArg constructs a boolean formula argument.
|
|
func newBoolFormulaArg(b bool) formulaArg {
|
|
var n float64
|
|
if b {
|
|
n = 1
|
|
}
|
|
return formulaArg{Type: ArgNumber, Number: n, Boolean: true}
|
|
}
|
|
|
|
// newErrorFormulaArg create an error formula argument of a given type with a
|
|
// specified error message.
|
|
func newErrorFormulaArg(formulaError, msg string) formulaArg {
|
|
return formulaArg{Type: ArgError, String: formulaError, Error: msg}
|
|
}
|
|
|
|
// newEmptyFormulaArg create an empty formula argument.
|
|
func newEmptyFormulaArg() formulaArg {
|
|
return formulaArg{Type: ArgEmpty}
|
|
}
|
|
|
|
// evalInfixExp evaluate syntax analysis by given infix expression after
|
|
// lexical analysis. Evaluate an infix expression containing formulas by
|
|
// stacks:
|
|
//
|
|
// opd - Operand
|
|
// opt - Operator
|
|
// opf - Operation formula
|
|
// opfd - Operand of the operation formula
|
|
// opft - Operator of the operation formula
|
|
// args - Arguments list of the operation formula
|
|
//
|
|
// TODO: handle subtypes: Nothing, Text, Logical, Error, Concatenation, Intersection, Union
|
|
func (f *File) evalInfixExp(ctx *calcContext, sheet, cell string, tokens []efp.Token) (formulaArg, error) {
|
|
var err error
|
|
opdStack, optStack, opfStack, opfdStack, opftStack, argsStack := NewStack(), NewStack(), NewStack(), NewStack(), NewStack(), NewStack()
|
|
var inArray, inArrayRow bool
|
|
for i := 0; i < len(tokens); i++ {
|
|
token := tokens[i]
|
|
|
|
// out of function stack
|
|
if opfStack.Len() == 0 {
|
|
if err = f.parseToken(ctx, sheet, token, opdStack, optStack); err != nil {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
}
|
|
|
|
// function start
|
|
if isFunctionStartToken(token) {
|
|
if token.TValue == "ARRAY" {
|
|
inArray = true
|
|
continue
|
|
}
|
|
if token.TValue == "ARRAYROW" {
|
|
inArrayRow = true
|
|
continue
|
|
}
|
|
opfStack.Push(token)
|
|
argsStack.Push(list.New().Init())
|
|
opftStack.Push(token) // to know which operators belong to a function use the function as a separator
|
|
continue
|
|
}
|
|
|
|
// in function stack, walk 2 token at once
|
|
if opfStack.Len() > 0 {
|
|
var nextToken efp.Token
|
|
if i+1 < len(tokens) {
|
|
nextToken = tokens[i+1]
|
|
}
|
|
|
|
// current token is args or range, skip next token, order required: parse reference first
|
|
if token.TSubType == efp.TokenSubTypeRange {
|
|
if opftStack.Peek().(efp.Token) != opfStack.Peek().(efp.Token) {
|
|
refTo := f.getDefinedNameRefTo(token.TValue, sheet)
|
|
if refTo != "" {
|
|
token.TValue = refTo
|
|
}
|
|
// parse reference: must reference at here
|
|
result, err := f.parseReference(ctx, sheet, token.TValue)
|
|
if err != nil {
|
|
return result, err
|
|
}
|
|
if result.Type == ArgError {
|
|
return result, errors.New(result.Error)
|
|
}
|
|
opfdStack.Push(result)
|
|
continue
|
|
}
|
|
if nextToken.TType == efp.TokenTypeArgument || nextToken.TType == efp.TokenTypeFunction {
|
|
// parse reference: reference or range at here
|
|
refTo := f.getDefinedNameRefTo(token.TValue, sheet)
|
|
if refTo != "" {
|
|
token.TValue = refTo
|
|
}
|
|
result, err := f.parseReference(ctx, sheet, token.TValue)
|
|
if err != nil {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
if result.Type == ArgUnknown {
|
|
return newEmptyFormulaArg(), errors.New(formulaErrorVALUE)
|
|
}
|
|
// when current token is range, next token is argument and opfdStack not empty,
|
|
// should push value to opfdStack and continue
|
|
if nextToken.TType == efp.TokenTypeArgument && !opfdStack.Empty() {
|
|
opfdStack.Push(result)
|
|
continue
|
|
}
|
|
argsStack.Peek().(*list.List).PushBack(result)
|
|
continue
|
|
}
|
|
}
|
|
|
|
if isEndParenthesesToken(token) && isBeginParenthesesToken(opftStack.Peek().(efp.Token)) {
|
|
if arg := argsStack.Peek().(*list.List).Back(); arg != nil {
|
|
opfdStack.Push(arg.Value.(formulaArg))
|
|
argsStack.Peek().(*list.List).Remove(arg)
|
|
}
|
|
}
|
|
|
|
// check current token is opft
|
|
if err = f.parseToken(ctx, sheet, token, opfdStack, opftStack); err != nil {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
|
|
// current token is arg
|
|
if token.TType == efp.TokenTypeArgument {
|
|
for opftStack.Peek().(efp.Token) != opfStack.Peek().(efp.Token) {
|
|
// calculate trigger
|
|
topOpt := opftStack.Peek().(efp.Token)
|
|
if err := calculate(opfdStack, topOpt); err != nil {
|
|
argsStack.Peek().(*list.List).PushFront(newErrorFormulaArg(formulaErrorVALUE, err.Error()))
|
|
}
|
|
opftStack.Pop()
|
|
}
|
|
if !opfdStack.Empty() {
|
|
argsStack.Peek().(*list.List).PushBack(opfdStack.Pop().(formulaArg))
|
|
}
|
|
continue
|
|
}
|
|
|
|
if inArrayRow && isOperand(token) {
|
|
continue
|
|
}
|
|
if inArrayRow && isFunctionStopToken(token) {
|
|
inArrayRow = false
|
|
continue
|
|
}
|
|
if inArray && isFunctionStopToken(token) {
|
|
argsStack.Peek().(*list.List).PushBack(opfdStack.Pop())
|
|
inArray = false
|
|
continue
|
|
}
|
|
if err = f.evalInfixExpFunc(ctx, sheet, cell, token, nextToken, opfStack, opdStack, opftStack, opfdStack, argsStack); err != nil {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
}
|
|
}
|
|
for optStack.Len() != 0 {
|
|
topOpt := optStack.Peek().(efp.Token)
|
|
if err = calculate(opdStack, topOpt); err != nil {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
optStack.Pop()
|
|
}
|
|
if opdStack.Len() == 0 {
|
|
return newEmptyFormulaArg(), ErrInvalidFormula
|
|
}
|
|
return opdStack.Peek().(formulaArg), err
|
|
}
|
|
|
|
// evalInfixExpFunc evaluate formula function in the infix expression.
|
|
func (f *File) evalInfixExpFunc(ctx *calcContext, sheet, cell string, token, nextToken efp.Token, opfStack, opdStack, opftStack, opfdStack, argsStack *Stack) error {
|
|
if !isFunctionStopToken(token) {
|
|
return nil
|
|
}
|
|
prepareEvalInfixExp(opfStack, opftStack, opfdStack, argsStack)
|
|
// call formula function to evaluate
|
|
arg := callFuncByName(&formulaFuncs{f: f, sheet: sheet, cell: cell, ctx: ctx}, strings.NewReplacer(
|
|
"_xlfn.", "", ".", "dot").Replace(opfStack.Peek().(efp.Token).TValue),
|
|
[]reflect.Value{reflect.ValueOf(argsStack.Peek().(*list.List))})
|
|
if arg.Type == ArgError && opfStack.Len() == 1 {
|
|
return errors.New(arg.Value())
|
|
}
|
|
argsStack.Pop()
|
|
opftStack.Pop() // remove current function separator
|
|
opfStack.Pop()
|
|
if opfStack.Len() > 0 { // still in function stack
|
|
if nextToken.TType == efp.TokenTypeOperatorInfix || (opftStack.Len() > 1 && opfdStack.Len() > 0) {
|
|
// mathematics calculate in formula function
|
|
opfdStack.Push(arg)
|
|
} else {
|
|
argsStack.Peek().(*list.List).PushBack(arg)
|
|
}
|
|
} else {
|
|
val := arg.Value()
|
|
if arg.Type == ArgMatrix && len(arg.Matrix) > 0 && len(arg.Matrix[0]) > 0 {
|
|
val = arg.Matrix[0][0].Value()
|
|
}
|
|
opdStack.Push(newStringFormulaArg(val))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// prepareEvalInfixExp check the token and stack state for formula function
|
|
// evaluate.
|
|
func prepareEvalInfixExp(opfStack, opftStack, opfdStack, argsStack *Stack) {
|
|
// current token is function stop
|
|
for opftStack.Peek().(efp.Token) != opfStack.Peek().(efp.Token) {
|
|
// calculate trigger
|
|
topOpt := opftStack.Peek().(efp.Token)
|
|
if err := calculate(opfdStack, topOpt); err != nil {
|
|
argsStack.Peek().(*list.List).PushBack(newErrorFormulaArg(err.Error(), err.Error()))
|
|
opftStack.Pop()
|
|
continue
|
|
}
|
|
opftStack.Pop()
|
|
}
|
|
argument := true
|
|
if opftStack.Len() > 2 && opfdStack.Len() == 1 {
|
|
topOpt := opftStack.Pop()
|
|
if opftStack.Peek().(efp.Token).TType == efp.TokenTypeOperatorInfix {
|
|
argument = false
|
|
}
|
|
opftStack.Push(topOpt)
|
|
}
|
|
// push opfd to args
|
|
if argument && opfdStack.Len() > 0 {
|
|
argsStack.Peek().(*list.List).PushBack(opfdStack.Pop().(formulaArg))
|
|
}
|
|
}
|
|
|
|
// calcPow evaluate exponentiation arithmetic operations.
|
|
func calcPow(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
lOpdVal := lOpd.ToNumber()
|
|
if lOpdVal.Type != ArgNumber {
|
|
return errors.New(lOpdVal.Value())
|
|
}
|
|
rOpdVal := rOpd.ToNumber()
|
|
if rOpdVal.Type != ArgNumber {
|
|
return errors.New(rOpdVal.Value())
|
|
}
|
|
opdStack.Push(newNumberFormulaArg(math.Pow(lOpdVal.Number, rOpdVal.Number)))
|
|
return nil
|
|
}
|
|
|
|
// calcEq evaluate equal arithmetic operations.
|
|
func calcEq(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
opdStack.Push(newBoolFormulaArg(rOpd.Value() == lOpd.Value()))
|
|
return nil
|
|
}
|
|
|
|
// calcNEq evaluate not equal arithmetic operations.
|
|
func calcNEq(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
opdStack.Push(newBoolFormulaArg(rOpd.Value() != lOpd.Value()))
|
|
return nil
|
|
}
|
|
|
|
// calcL evaluate less than arithmetic operations.
|
|
func calcL(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(lOpd.Number < rOpd.Number))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(strings.Compare(lOpd.Value(), rOpd.Value()) == -1))
|
|
}
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(false))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(true))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// calcLe evaluate less than or equal arithmetic operations.
|
|
func calcLe(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(lOpd.Number <= rOpd.Number))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(strings.Compare(lOpd.Value(), rOpd.Value()) != 1))
|
|
}
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(false))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(true))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// calcG evaluate greater than arithmetic operations.
|
|
func calcG(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(lOpd.Number > rOpd.Number))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(strings.Compare(lOpd.Value(), rOpd.Value()) == 1))
|
|
}
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(true))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(false))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// calcGe evaluate greater than or equal arithmetic operations.
|
|
func calcGe(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(lOpd.Number >= rOpd.Number))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(strings.Compare(lOpd.Value(), rOpd.Value()) != -1))
|
|
}
|
|
if rOpd.Type == ArgNumber && lOpd.Type == ArgString {
|
|
opdStack.Push(newBoolFormulaArg(true))
|
|
}
|
|
if rOpd.Type == ArgString && lOpd.Type == ArgNumber {
|
|
opdStack.Push(newBoolFormulaArg(false))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// calcSplice evaluate splice '&' operations.
|
|
func calcSplice(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
opdStack.Push(newStringFormulaArg(lOpd.Value() + rOpd.Value()))
|
|
return nil
|
|
}
|
|
|
|
// calcAdd evaluate addition arithmetic operations.
|
|
func calcAdd(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
lOpdVal := lOpd.ToNumber()
|
|
if lOpdVal.Type != ArgNumber {
|
|
return errors.New(lOpdVal.Value())
|
|
}
|
|
rOpdVal := rOpd.ToNumber()
|
|
if rOpdVal.Type != ArgNumber {
|
|
return errors.New(rOpdVal.Value())
|
|
}
|
|
opdStack.Push(newNumberFormulaArg(lOpdVal.Number + rOpdVal.Number))
|
|
return nil
|
|
}
|
|
|
|
// calcSubtract evaluate subtraction arithmetic operations.
|
|
func calcSubtract(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
lOpdVal := lOpd.ToNumber()
|
|
if lOpdVal.Type != ArgNumber {
|
|
return errors.New(lOpdVal.Value())
|
|
}
|
|
rOpdVal := rOpd.ToNumber()
|
|
if rOpdVal.Type != ArgNumber {
|
|
return errors.New(rOpdVal.Value())
|
|
}
|
|
opdStack.Push(newNumberFormulaArg(lOpdVal.Number - rOpdVal.Number))
|
|
return nil
|
|
}
|
|
|
|
// calcMultiply evaluate multiplication arithmetic operations.
|
|
func calcMultiply(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
lOpdVal := lOpd.ToNumber()
|
|
if lOpdVal.Type != ArgNumber {
|
|
return errors.New(lOpdVal.Value())
|
|
}
|
|
rOpdVal := rOpd.ToNumber()
|
|
if rOpdVal.Type != ArgNumber {
|
|
return errors.New(rOpdVal.Value())
|
|
}
|
|
opdStack.Push(newNumberFormulaArg(lOpdVal.Number * rOpdVal.Number))
|
|
return nil
|
|
}
|
|
|
|
// calcDiv evaluate division arithmetic operations.
|
|
func calcDiv(rOpd, lOpd formulaArg, opdStack *Stack) error {
|
|
lOpdVal := lOpd.ToNumber()
|
|
if lOpdVal.Type != ArgNumber {
|
|
return errors.New(lOpdVal.Value())
|
|
}
|
|
rOpdVal := rOpd.ToNumber()
|
|
if rOpdVal.Type != ArgNumber {
|
|
return errors.New(rOpdVal.Value())
|
|
}
|
|
if rOpdVal.Number == 0 {
|
|
return errors.New(formulaErrorDIV)
|
|
}
|
|
opdStack.Push(newNumberFormulaArg(lOpdVal.Number / rOpdVal.Number))
|
|
return nil
|
|
}
|
|
|
|
// calculate evaluate basic arithmetic operations.
|
|
func calculate(opdStack *Stack, opt efp.Token) error {
|
|
if opt.TValue == "-" && opt.TType == efp.TokenTypeOperatorPrefix {
|
|
if opdStack.Len() < 1 {
|
|
return ErrInvalidFormula
|
|
}
|
|
opd := opdStack.Pop().(formulaArg)
|
|
opdStack.Push(newNumberFormulaArg(0 - opd.ToNumber().Number))
|
|
}
|
|
if opt.TValue == "-" && opt.TType == efp.TokenTypeOperatorInfix {
|
|
if opdStack.Len() < 2 {
|
|
return ErrInvalidFormula
|
|
}
|
|
rOpd := opdStack.Pop().(formulaArg)
|
|
lOpd := opdStack.Pop().(formulaArg)
|
|
if err := calcSubtract(rOpd, lOpd, opdStack); err != nil {
|
|
return err
|
|
}
|
|
}
|
|
tokenCalcFunc := map[string]func(rOpd, lOpd formulaArg, opdStack *Stack) error{
|
|
"^": calcPow,
|
|
"*": calcMultiply,
|
|
"/": calcDiv,
|
|
"+": calcAdd,
|
|
"=": calcEq,
|
|
"<>": calcNEq,
|
|
"<": calcL,
|
|
"<=": calcLe,
|
|
">": calcG,
|
|
">=": calcGe,
|
|
"&": calcSplice,
|
|
}
|
|
fn, ok := tokenCalcFunc[opt.TValue]
|
|
if ok {
|
|
if opdStack.Len() < 2 {
|
|
return ErrInvalidFormula
|
|
}
|
|
rOpd := opdStack.Pop().(formulaArg)
|
|
lOpd := opdStack.Pop().(formulaArg)
|
|
if rOpd.Type == ArgError {
|
|
return errors.New(rOpd.Value())
|
|
}
|
|
if lOpd.Type == ArgError {
|
|
return errors.New(lOpd.Value())
|
|
}
|
|
if err := fn(rOpd, lOpd, opdStack); err != nil {
|
|
return err
|
|
}
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// parseOperatorPrefixToken parse operator prefix token.
|
|
func (f *File) parseOperatorPrefixToken(optStack, opdStack *Stack, token efp.Token) (err error) {
|
|
if optStack.Len() == 0 {
|
|
optStack.Push(token)
|
|
return
|
|
}
|
|
tokenPriority := getPriority(token)
|
|
topOpt := optStack.Peek().(efp.Token)
|
|
topOptPriority := getPriority(topOpt)
|
|
if tokenPriority > topOptPriority {
|
|
optStack.Push(token)
|
|
return
|
|
}
|
|
for tokenPriority <= topOptPriority {
|
|
optStack.Pop()
|
|
if err = calculate(opdStack, topOpt); err != nil {
|
|
return
|
|
}
|
|
if optStack.Len() > 0 {
|
|
topOpt = optStack.Peek().(efp.Token)
|
|
topOptPriority = getPriority(topOpt)
|
|
continue
|
|
}
|
|
break
|
|
}
|
|
optStack.Push(token)
|
|
return
|
|
}
|
|
|
|
// isFunctionStartToken determine if the token is function start.
|
|
func isFunctionStartToken(token efp.Token) bool {
|
|
return token.TType == efp.TokenTypeFunction && token.TSubType == efp.TokenSubTypeStart
|
|
}
|
|
|
|
// isFunctionStopToken determine if the token is function stop.
|
|
func isFunctionStopToken(token efp.Token) bool {
|
|
return token.TType == efp.TokenTypeFunction && token.TSubType == efp.TokenSubTypeStop
|
|
}
|
|
|
|
// isBeginParenthesesToken determine if the token is begin parentheses: (.
|
|
func isBeginParenthesesToken(token efp.Token) bool {
|
|
return token.TType == efp.TokenTypeSubexpression && token.TSubType == efp.TokenSubTypeStart
|
|
}
|
|
|
|
// isEndParenthesesToken determine if the token is end parentheses: ).
|
|
func isEndParenthesesToken(token efp.Token) bool {
|
|
return token.TType == efp.TokenTypeSubexpression && token.TSubType == efp.TokenSubTypeStop
|
|
}
|
|
|
|
// isOperatorPrefixToken determine if the token is parse operator prefix
|
|
// token.
|
|
func isOperatorPrefixToken(token efp.Token) bool {
|
|
_, ok := tokenPriority[token.TValue]
|
|
return (token.TValue == "-" && token.TType == efp.TokenTypeOperatorPrefix) || (ok && token.TType == efp.TokenTypeOperatorInfix)
|
|
}
|
|
|
|
// isOperand determine if the token is parse operand.
|
|
func isOperand(token efp.Token) bool {
|
|
return token.TType == efp.TokenTypeOperand && (token.TSubType == efp.TokenSubTypeNumber || token.TSubType == efp.TokenSubTypeText || token.TSubType == efp.TokenSubTypeLogical)
|
|
}
|
|
|
|
// tokenToFormulaArg create a formula argument by given token.
|
|
func tokenToFormulaArg(token efp.Token) formulaArg {
|
|
switch token.TSubType {
|
|
case efp.TokenSubTypeLogical:
|
|
return newBoolFormulaArg(strings.EqualFold(token.TValue, "TRUE"))
|
|
case efp.TokenSubTypeNumber:
|
|
num, _ := strconv.ParseFloat(token.TValue, 64)
|
|
return newNumberFormulaArg(num)
|
|
default:
|
|
return newStringFormulaArg(token.TValue)
|
|
}
|
|
}
|
|
|
|
// formulaArgToToken create a token by given formula argument.
|
|
func formulaArgToToken(arg formulaArg) efp.Token {
|
|
switch arg.Type {
|
|
case ArgNumber:
|
|
if arg.Boolean {
|
|
return efp.Token{TValue: arg.Value(), TType: efp.TokenTypeOperand, TSubType: efp.TokenSubTypeLogical}
|
|
}
|
|
return efp.Token{TValue: arg.Value(), TType: efp.TokenTypeOperand, TSubType: efp.TokenSubTypeNumber}
|
|
default:
|
|
return efp.Token{TValue: arg.Value(), TType: efp.TokenTypeOperand, TSubType: efp.TokenSubTypeText}
|
|
}
|
|
}
|
|
|
|
// parseToken parse basic arithmetic operator priority and evaluate based on
|
|
// operators and operands.
|
|
func (f *File) parseToken(ctx *calcContext, sheet string, token efp.Token, opdStack, optStack *Stack) error {
|
|
// parse reference: must reference at here
|
|
if token.TSubType == efp.TokenSubTypeRange {
|
|
refTo := f.getDefinedNameRefTo(token.TValue, sheet)
|
|
if refTo != "" {
|
|
token.TValue = refTo
|
|
}
|
|
result, err := f.parseReference(ctx, sheet, token.TValue)
|
|
if err != nil {
|
|
return errors.New(formulaErrorNAME)
|
|
}
|
|
token = formulaArgToToken(result)
|
|
}
|
|
if isOperatorPrefixToken(token) {
|
|
if err := f.parseOperatorPrefixToken(optStack, opdStack, token); err != nil {
|
|
return err
|
|
}
|
|
}
|
|
if isBeginParenthesesToken(token) { // (
|
|
optStack.Push(token)
|
|
}
|
|
if isEndParenthesesToken(token) { // )
|
|
for !isBeginParenthesesToken(optStack.Peek().(efp.Token)) { // != (
|
|
topOpt := optStack.Peek().(efp.Token)
|
|
if err := calculate(opdStack, topOpt); err != nil {
|
|
return err
|
|
}
|
|
optStack.Pop()
|
|
}
|
|
optStack.Pop()
|
|
}
|
|
if token.TType == efp.TokenTypeOperatorPostfix && !opdStack.Empty() {
|
|
topOpd := opdStack.Pop().(formulaArg)
|
|
opdStack.Push(newNumberFormulaArg(topOpd.Number / 100))
|
|
}
|
|
// opd
|
|
if isOperand(token) {
|
|
opdStack.Push(tokenToFormulaArg(token))
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// parseReference parse reference and extract values by given reference
|
|
// characters and default sheet name.
|
|
func (f *File) parseReference(ctx *calcContext, sheet, reference string) (arg formulaArg, err error) {
|
|
reference = strings.ReplaceAll(reference, "$", "")
|
|
refs, cellRanges, cellRefs := list.New(), list.New(), list.New()
|
|
for _, ref := range strings.Split(reference, ":") {
|
|
tokens := strings.Split(ref, "!")
|
|
cr := cellRef{}
|
|
if len(tokens) == 2 { // have a worksheet name
|
|
cr.Sheet = tokens[0]
|
|
// cast to cell reference
|
|
if cr.Col, cr.Row, err = CellNameToCoordinates(tokens[1]); err != nil {
|
|
// cast to column
|
|
if cr.Col, err = ColumnNameToNumber(tokens[1]); err != nil {
|
|
// cast to row
|
|
if cr.Row, err = strconv.Atoi(tokens[1]); err != nil {
|
|
err = newInvalidColumnNameError(tokens[1])
|
|
return
|
|
}
|
|
cr.Col = MaxColumns
|
|
}
|
|
}
|
|
if refs.Len() > 0 {
|
|
e := refs.Back()
|
|
cellRefs.PushBack(e.Value.(cellRef))
|
|
refs.Remove(e)
|
|
}
|
|
refs.PushBack(cr)
|
|
continue
|
|
}
|
|
// cast to cell reference
|
|
if cr.Col, cr.Row, err = CellNameToCoordinates(tokens[0]); err != nil {
|
|
// cast to column
|
|
if cr.Col, err = ColumnNameToNumber(tokens[0]); err != nil {
|
|
// cast to row
|
|
if cr.Row, err = strconv.Atoi(tokens[0]); err != nil {
|
|
err = newInvalidColumnNameError(tokens[0])
|
|
return
|
|
}
|
|
cr.Col = MaxColumns
|
|
}
|
|
cellRanges.PushBack(cellRange{
|
|
From: cellRef{Sheet: sheet, Col: cr.Col, Row: 1},
|
|
To: cellRef{Sheet: sheet, Col: cr.Col, Row: TotalRows},
|
|
})
|
|
cellRefs.Init()
|
|
arg, err = f.rangeResolver(ctx, cellRefs, cellRanges)
|
|
return
|
|
}
|
|
e := refs.Back()
|
|
if e == nil {
|
|
cr.Sheet = sheet
|
|
refs.PushBack(cr)
|
|
continue
|
|
}
|
|
cellRanges.PushBack(cellRange{
|
|
From: e.Value.(cellRef),
|
|
To: cr,
|
|
})
|
|
refs.Remove(e)
|
|
}
|
|
if refs.Len() > 0 {
|
|
e := refs.Back()
|
|
cellRefs.PushBack(e.Value.(cellRef))
|
|
refs.Remove(e)
|
|
}
|
|
arg, err = f.rangeResolver(ctx, cellRefs, cellRanges)
|
|
return
|
|
}
|
|
|
|
// prepareValueRange prepare value range.
|
|
func prepareValueRange(cr cellRange, valueRange []int) {
|
|
if cr.From.Row < valueRange[0] || valueRange[0] == 0 {
|
|
valueRange[0] = cr.From.Row
|
|
}
|
|
if cr.From.Col < valueRange[2] || valueRange[2] == 0 {
|
|
valueRange[2] = cr.From.Col
|
|
}
|
|
if cr.To.Row > valueRange[1] || valueRange[1] == 0 {
|
|
valueRange[1] = cr.To.Row
|
|
}
|
|
if cr.To.Col > valueRange[3] || valueRange[3] == 0 {
|
|
valueRange[3] = cr.To.Col
|
|
}
|
|
}
|
|
|
|
// prepareValueRef prepare value reference.
|
|
func prepareValueRef(cr cellRef, valueRange []int) {
|
|
if cr.Row < valueRange[0] || valueRange[0] == 0 {
|
|
valueRange[0] = cr.Row
|
|
}
|
|
if cr.Col < valueRange[2] || valueRange[2] == 0 {
|
|
valueRange[2] = cr.Col
|
|
}
|
|
if cr.Row > valueRange[1] || valueRange[1] == 0 {
|
|
valueRange[1] = cr.Row
|
|
}
|
|
if cr.Col > valueRange[3] || valueRange[3] == 0 {
|
|
valueRange[3] = cr.Col
|
|
}
|
|
}
|
|
|
|
// cellResolver calc cell value by given worksheet name, cell reference and context.
|
|
func (f *File) cellResolver(ctx *calcContext, sheet, cell string) (formulaArg, error) {
|
|
var (
|
|
arg formulaArg
|
|
value string
|
|
err error
|
|
)
|
|
if value, err = f.GetCellValue(sheet, cell, Options{RawCellValue: true}); err != nil {
|
|
return arg, err
|
|
}
|
|
arg = newStringFormulaArg(value)
|
|
cellType, _ := f.GetCellType(sheet, cell)
|
|
switch cellType {
|
|
case CellTypeBool:
|
|
return arg.ToBool(), err
|
|
case CellTypeNumber, CellTypeUnset:
|
|
if arg.Value() == "" {
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
return arg.ToNumber(), err
|
|
case CellTypeInlineString, CellTypeSharedString:
|
|
return arg, err
|
|
case CellTypeFormula:
|
|
ref := fmt.Sprintf("%s!%s", sheet, cell)
|
|
if ctx.entry != ref {
|
|
ctx.Lock()
|
|
if ctx.iterations[ref] <= ctx.maxCalcIterations {
|
|
ctx.iterations[ref]++
|
|
ctx.Unlock()
|
|
arg, _ = f.calcCellValue(ctx, sheet, cell)
|
|
ctx.iterationsCache[ref] = arg
|
|
return arg, nil
|
|
}
|
|
ctx.Unlock()
|
|
return ctx.iterationsCache[ref], nil
|
|
}
|
|
fallthrough
|
|
default:
|
|
return newEmptyFormulaArg(), err
|
|
}
|
|
}
|
|
|
|
// rangeResolver extract value as string from given reference and range list.
|
|
// This function will not ignore the empty cell. For example, A1:A2:A2:B3 will
|
|
// be reference A1:B3.
|
|
func (f *File) rangeResolver(ctx *calcContext, cellRefs, cellRanges *list.List) (arg formulaArg, err error) {
|
|
arg.cellRefs, arg.cellRanges = cellRefs, cellRanges
|
|
// value range order: from row, to row, from column, to column
|
|
valueRange := []int{0, 0, 0, 0}
|
|
var sheet string
|
|
// prepare value range
|
|
for temp := cellRanges.Front(); temp != nil; temp = temp.Next() {
|
|
cr := temp.Value.(cellRange)
|
|
if cr.From.Sheet != cr.To.Sheet {
|
|
err = errors.New(formulaErrorVALUE)
|
|
}
|
|
rng := []int{cr.From.Col, cr.From.Row, cr.To.Col, cr.To.Row}
|
|
_ = sortCoordinates(rng)
|
|
cr.From.Col, cr.From.Row, cr.To.Col, cr.To.Row = rng[0], rng[1], rng[2], rng[3]
|
|
prepareValueRange(cr, valueRange)
|
|
if cr.From.Sheet != "" {
|
|
sheet = cr.From.Sheet
|
|
}
|
|
}
|
|
for temp := cellRefs.Front(); temp != nil; temp = temp.Next() {
|
|
cr := temp.Value.(cellRef)
|
|
if cr.Sheet != "" {
|
|
sheet = cr.Sheet
|
|
}
|
|
prepareValueRef(cr, valueRange)
|
|
}
|
|
// extract value from ranges
|
|
if cellRanges.Len() > 0 {
|
|
arg.Type = ArgMatrix
|
|
for row := valueRange[0]; row <= valueRange[1]; row++ {
|
|
var matrixRow []formulaArg
|
|
for col := valueRange[2]; col <= valueRange[3]; col++ {
|
|
var cell string
|
|
var value formulaArg
|
|
if cell, err = CoordinatesToCellName(col, row); err != nil {
|
|
return
|
|
}
|
|
if value, err = f.cellResolver(ctx, sheet, cell); err != nil {
|
|
return
|
|
}
|
|
matrixRow = append(matrixRow, value)
|
|
}
|
|
arg.Matrix = append(arg.Matrix, matrixRow)
|
|
}
|
|
return
|
|
}
|
|
// extract value from references
|
|
for temp := cellRefs.Front(); temp != nil; temp = temp.Next() {
|
|
cr := temp.Value.(cellRef)
|
|
var cell string
|
|
if cell, err = CoordinatesToCellName(cr.Col, cr.Row); err != nil {
|
|
return
|
|
}
|
|
if arg, err = f.cellResolver(ctx, cr.Sheet, cell); err != nil {
|
|
return
|
|
}
|
|
arg.cellRefs, arg.cellRanges = cellRefs, cellRanges
|
|
}
|
|
return
|
|
}
|
|
|
|
// callFuncByName calls the no error or only error return function with
|
|
// reflect by given receiver, name and parameters.
|
|
func callFuncByName(receiver interface{}, name string, params []reflect.Value) (arg formulaArg) {
|
|
function := reflect.ValueOf(receiver).MethodByName(name)
|
|
if function.IsValid() {
|
|
rt := function.Call(params)
|
|
if len(rt) == 0 {
|
|
return
|
|
}
|
|
arg = rt[0].Interface().(formulaArg)
|
|
return
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("not support %s function", name))
|
|
}
|
|
|
|
// formulaCriteriaParser parse formula criteria.
|
|
func formulaCriteriaParser(exp string) (fc *formulaCriteria) {
|
|
fc = &formulaCriteria{}
|
|
if exp == "" {
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^(\d+)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaEq, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^=(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaEq, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^<>(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaNe, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^<=(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaLe, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^>=(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaGe, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^<(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaL, match[1]
|
|
return
|
|
}
|
|
if match := regexp.MustCompile(`^>(.*)$`).FindStringSubmatch(exp); len(match) > 1 {
|
|
fc.Type, fc.Condition = criteriaG, match[1]
|
|
return
|
|
}
|
|
if strings.Contains(exp, "?") {
|
|
exp = strings.ReplaceAll(exp, "?", ".")
|
|
}
|
|
if strings.Contains(exp, "*") {
|
|
exp = strings.ReplaceAll(exp, "*", ".*")
|
|
}
|
|
fc.Type, fc.Condition = criteriaRegexp, exp
|
|
return
|
|
}
|
|
|
|
// formulaCriteriaEval evaluate formula criteria expression.
|
|
func formulaCriteriaEval(val string, criteria *formulaCriteria) (result bool, err error) {
|
|
var value, expected float64
|
|
var e error
|
|
prepareValue := func(val, cond string) (value float64, expected float64, err error) {
|
|
percentile := 1.0
|
|
if strings.HasSuffix(cond, "%") {
|
|
cond = strings.TrimSuffix(cond, "%")
|
|
percentile /= 100
|
|
}
|
|
if value, err = strconv.ParseFloat(val, 64); err != nil {
|
|
return
|
|
}
|
|
if expected, err = strconv.ParseFloat(cond, 64); err != nil {
|
|
return
|
|
}
|
|
expected *= percentile
|
|
return
|
|
}
|
|
switch criteria.Type {
|
|
case criteriaEq:
|
|
return val == criteria.Condition, err
|
|
case criteriaLe:
|
|
value, expected, e = prepareValue(val, criteria.Condition)
|
|
return value <= expected && e == nil, err
|
|
case criteriaGe:
|
|
value, expected, e = prepareValue(val, criteria.Condition)
|
|
return value >= expected && e == nil, err
|
|
case criteriaNe:
|
|
return val != criteria.Condition, err
|
|
case criteriaL:
|
|
value, expected, e = prepareValue(val, criteria.Condition)
|
|
return value < expected && e == nil, err
|
|
case criteriaG:
|
|
value, expected, e = prepareValue(val, criteria.Condition)
|
|
return value > expected && e == nil, err
|
|
case criteriaRegexp:
|
|
return regexp.MatchString(criteria.Condition, val)
|
|
}
|
|
return
|
|
}
|
|
|
|
// Engineering Functions
|
|
|
|
// BESSELI function the modified Bessel function, which is equivalent to the
|
|
// Bessel function evaluated for purely imaginary arguments. The syntax of
|
|
// the Besseli function is:
|
|
//
|
|
// BESSELI(x,n)
|
|
func (fn *formulaFuncs) BESSELI(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BESSELI requires 2 numeric arguments")
|
|
}
|
|
return fn.bassel(argsList, true)
|
|
}
|
|
|
|
// BESSELJ function returns the Bessel function, Jn(x), for a specified order
|
|
// and value of x. The syntax of the function is:
|
|
//
|
|
// BESSELJ(x,n)
|
|
func (fn *formulaFuncs) BESSELJ(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BESSELJ requires 2 numeric arguments")
|
|
}
|
|
return fn.bassel(argsList, false)
|
|
}
|
|
|
|
// bassel is an implementation of the formula functions BESSELI and BESSELJ.
|
|
func (fn *formulaFuncs) bassel(argsList *list.List, modfied bool) formulaArg {
|
|
x, n := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Back().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if n.Type != ArgNumber {
|
|
return n
|
|
}
|
|
max, x1 := 100, x.Number*0.5
|
|
x2 := x1 * x1
|
|
x1 = math.Pow(x1, n.Number)
|
|
n1, n2, n3, n4, add := fact(n.Number), 1.0, 0.0, n.Number, false
|
|
result := x1 / n1
|
|
t := result * 0.9
|
|
for result != t && max != 0 {
|
|
x1 *= x2
|
|
n3++
|
|
n1 *= n3
|
|
n4++
|
|
n2 *= n4
|
|
t = result
|
|
r := x1 / n1 / n2
|
|
if modfied || add {
|
|
result += r
|
|
} else {
|
|
result -= r
|
|
}
|
|
max--
|
|
add = !add
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// BESSELK function calculates the modified Bessel functions, Kn(x), which are
|
|
// also known as the hyperbolic Bessel Functions. These are the equivalent of
|
|
// the Bessel functions, evaluated for purely imaginary arguments. The syntax
|
|
// of the function is:
|
|
//
|
|
// BESSELK(x,n)
|
|
func (fn *formulaFuncs) BESSELK(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BESSELK requires 2 numeric arguments")
|
|
}
|
|
x, n := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Back().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if n.Type != ArgNumber {
|
|
return n
|
|
}
|
|
if x.Number <= 0 || n.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
var result float64
|
|
switch math.Floor(n.Number) {
|
|
case 0:
|
|
result = fn.besselK0(x)
|
|
case 1:
|
|
result = fn.besselK1(x)
|
|
default:
|
|
result = fn.besselK2(x, n)
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// besselK0 is an implementation of the formula function BESSELK.
|
|
func (fn *formulaFuncs) besselK0(x formulaArg) float64 {
|
|
var y float64
|
|
if x.Number <= 2 {
|
|
n2 := x.Number * 0.5
|
|
y = n2 * n2
|
|
args := list.New()
|
|
args.PushBack(x)
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
return -math.Log(n2)*fn.BESSELI(args).Number +
|
|
(-0.57721566 + y*(0.42278420+y*(0.23069756+y*(0.3488590e-1+y*(0.262698e-2+y*
|
|
(0.10750e-3+y*0.74e-5))))))
|
|
}
|
|
y = 2 / x.Number
|
|
return math.Exp(-x.Number) / math.Sqrt(x.Number) *
|
|
(1.25331414 + y*(-0.7832358e-1+y*(0.2189568e-1+y*(-0.1062446e-1+y*
|
|
(0.587872e-2+y*(-0.251540e-2+y*0.53208e-3))))))
|
|
}
|
|
|
|
// besselK1 is an implementation of the formula function BESSELK.
|
|
func (fn *formulaFuncs) besselK1(x formulaArg) float64 {
|
|
var n2, y float64
|
|
if x.Number <= 2 {
|
|
n2 = x.Number * 0.5
|
|
y = n2 * n2
|
|
args := list.New()
|
|
args.PushBack(x)
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
return math.Log(n2)*fn.BESSELI(args).Number +
|
|
(1+y*(0.15443144+y*(-0.67278579+y*(-0.18156897+y*(-0.1919402e-1+y*(-0.110404e-2+y*(-0.4686e-4)))))))/x.Number
|
|
}
|
|
y = 2 / x.Number
|
|
return math.Exp(-x.Number) / math.Sqrt(x.Number) *
|
|
(1.25331414 + y*(0.23498619+y*(-0.3655620e-1+y*(0.1504268e-1+y*(-0.780353e-2+y*
|
|
(0.325614e-2+y*(-0.68245e-3)))))))
|
|
}
|
|
|
|
// besselK2 is an implementation of the formula function BESSELK.
|
|
func (fn *formulaFuncs) besselK2(x, n formulaArg) float64 {
|
|
tox, bkm, bk, bkp := 2/x.Number, fn.besselK0(x), fn.besselK1(x), 0.0
|
|
for i := 1.0; i < n.Number; i++ {
|
|
bkp = bkm + i*tox*bk
|
|
bkm = bk
|
|
bk = bkp
|
|
}
|
|
return bk
|
|
}
|
|
|
|
// BESSELY function returns the Bessel function, Yn(x), (also known as the
|
|
// Weber function or the Neumann function), for a specified order and value
|
|
// of x. The syntax of the function is:
|
|
//
|
|
// BESSELY(x,n)
|
|
func (fn *formulaFuncs) BESSELY(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BESSELY requires 2 numeric arguments")
|
|
}
|
|
x, n := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Back().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if n.Type != ArgNumber {
|
|
return n
|
|
}
|
|
if x.Number <= 0 || n.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
var result float64
|
|
switch math.Floor(n.Number) {
|
|
case 0:
|
|
result = fn.besselY0(x)
|
|
case 1:
|
|
result = fn.besselY1(x)
|
|
default:
|
|
result = fn.besselY2(x, n)
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// besselY0 is an implementation of the formula function BESSELY.
|
|
func (fn *formulaFuncs) besselY0(x formulaArg) float64 {
|
|
var y float64
|
|
if x.Number < 8 {
|
|
y = x.Number * x.Number
|
|
f1 := -2957821389.0 + y*(7062834065.0+y*(-512359803.6+y*(10879881.29+y*
|
|
(-86327.92757+y*228.4622733))))
|
|
f2 := 40076544269.0 + y*(745249964.8+y*(7189466.438+y*
|
|
(47447.26470+y*(226.1030244+y))))
|
|
args := list.New()
|
|
args.PushBack(x)
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
return f1/f2 + 0.636619772*fn.BESSELJ(args).Number*math.Log(x.Number)
|
|
}
|
|
z := 8.0 / x.Number
|
|
y = z * z
|
|
xx := x.Number - 0.785398164
|
|
f1 := 1 + y*(-0.1098628627e-2+y*(0.2734510407e-4+y*(-0.2073370639e-5+y*0.2093887211e-6)))
|
|
f2 := -0.1562499995e-1 + y*(0.1430488765e-3+y*(-0.6911147651e-5+y*(0.7621095161e-6+y*
|
|
(-0.934945152e-7))))
|
|
return math.Sqrt(0.636619772/x.Number) * (math.Sin(xx)*f1 + z*math.Cos(xx)*f2)
|
|
}
|
|
|
|
// besselY1 is an implementation of the formula function BESSELY.
|
|
func (fn *formulaFuncs) besselY1(x formulaArg) float64 {
|
|
if x.Number < 8 {
|
|
y := x.Number * x.Number
|
|
f1 := x.Number * (-0.4900604943e13 + y*(0.1275274390e13+y*(-0.5153438139e11+y*
|
|
(0.7349264551e9+y*(-0.4237922726e7+y*0.8511937935e4)))))
|
|
f2 := 0.2499580570e14 + y*(0.4244419664e12+y*(0.3733650367e10+y*(0.2245904002e8+y*
|
|
(0.1020426050e6+y*(0.3549632885e3+y)))))
|
|
args := list.New()
|
|
args.PushBack(x)
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
return f1/f2 + 0.636619772*(fn.BESSELJ(args).Number*math.Log(x.Number)-1/x.Number)
|
|
}
|
|
return math.Sqrt(0.636619772/x.Number) * math.Sin(x.Number-2.356194491)
|
|
}
|
|
|
|
// besselY2 is an implementation of the formula function BESSELY.
|
|
func (fn *formulaFuncs) besselY2(x, n formulaArg) float64 {
|
|
tox, bym, by, byp := 2/x.Number, fn.besselY0(x), fn.besselY1(x), 0.0
|
|
for i := 1.0; i < n.Number; i++ {
|
|
byp = i*tox*by - bym
|
|
bym = by
|
|
by = byp
|
|
}
|
|
return by
|
|
}
|
|
|
|
// BIN2DEC function converts a Binary (a base-2 number) into a decimal number.
|
|
// The syntax of the function is:
|
|
//
|
|
// BIN2DEC(number)
|
|
func (fn *formulaFuncs) BIN2DEC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BIN2DEC requires 1 numeric argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
return fn.bin2dec(token.Value())
|
|
}
|
|
|
|
// BIN2HEX function converts a Binary (Base 2) number into a Hexadecimal
|
|
// (Base 16) number. The syntax of the function is:
|
|
//
|
|
// BIN2HEX(number,[places])
|
|
func (fn *formulaFuncs) BIN2HEX(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BIN2HEX requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BIN2HEX allows at most 2 arguments")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
decimal, newList := fn.bin2dec(token.Value()), list.New()
|
|
if decimal.Type != ArgNumber {
|
|
return decimal
|
|
}
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("BIN2HEX", newList)
|
|
}
|
|
|
|
// BIN2OCT function converts a Binary (Base 2) number into an Octal (Base 8)
|
|
// number. The syntax of the function is:
|
|
//
|
|
// BIN2OCT(number,[places])
|
|
func (fn *formulaFuncs) BIN2OCT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BIN2OCT requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BIN2OCT allows at most 2 arguments")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
decimal, newList := fn.bin2dec(token.Value()), list.New()
|
|
if decimal.Type != ArgNumber {
|
|
return decimal
|
|
}
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("BIN2OCT", newList)
|
|
}
|
|
|
|
// bin2dec is an implementation of the formula function BIN2DEC.
|
|
func (fn *formulaFuncs) bin2dec(number string) formulaArg {
|
|
decimal, length := 0.0, len(number)
|
|
for i := length; i > 0; i-- {
|
|
s := string(number[length-i])
|
|
if i == 10 && s == "1" {
|
|
decimal += math.Pow(-2.0, float64(i-1))
|
|
continue
|
|
}
|
|
if s == "1" {
|
|
decimal += math.Pow(2.0, float64(i-1))
|
|
continue
|
|
}
|
|
if s != "0" {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return newNumberFormulaArg(decimal)
|
|
}
|
|
|
|
// BITAND function returns the bitwise 'AND' for two supplied integers. The
|
|
// syntax of the function is:
|
|
//
|
|
// BITAND(number1,number2)
|
|
func (fn *formulaFuncs) BITAND(argsList *list.List) formulaArg {
|
|
return fn.bitwise("BITAND", argsList)
|
|
}
|
|
|
|
// BITLSHIFT function returns a supplied integer, shifted left by a specified
|
|
// number of bits. The syntax of the function is:
|
|
//
|
|
// BITLSHIFT(number1,shift_amount)
|
|
func (fn *formulaFuncs) BITLSHIFT(argsList *list.List) formulaArg {
|
|
return fn.bitwise("BITLSHIFT", argsList)
|
|
}
|
|
|
|
// BITOR function returns the bitwise 'OR' for two supplied integers. The
|
|
// syntax of the function is:
|
|
//
|
|
// BITOR(number1,number2)
|
|
func (fn *formulaFuncs) BITOR(argsList *list.List) formulaArg {
|
|
return fn.bitwise("BITOR", argsList)
|
|
}
|
|
|
|
// BITRSHIFT function returns a supplied integer, shifted right by a specified
|
|
// number of bits. The syntax of the function is:
|
|
//
|
|
// BITRSHIFT(number1,shift_amount)
|
|
func (fn *formulaFuncs) BITRSHIFT(argsList *list.List) formulaArg {
|
|
return fn.bitwise("BITRSHIFT", argsList)
|
|
}
|
|
|
|
// BITXOR function returns the bitwise 'XOR' (exclusive 'OR') for two supplied
|
|
// integers. The syntax of the function is:
|
|
//
|
|
// BITXOR(number1,number2)
|
|
func (fn *formulaFuncs) BITXOR(argsList *list.List) formulaArg {
|
|
return fn.bitwise("BITXOR", argsList)
|
|
}
|
|
|
|
// bitwise is an implementation of the formula functions BITAND, BITLSHIFT,
|
|
// BITOR, BITRSHIFT and BITXOR.
|
|
func (fn *formulaFuncs) bitwise(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 numeric arguments", name))
|
|
}
|
|
num1, num2 := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Back().Value.(formulaArg).ToNumber()
|
|
if num1.Type != ArgNumber || num2.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
max := math.Pow(2, 48) - 1
|
|
if num1.Number < 0 || num1.Number > max || num2.Number < 0 || num2.Number > max {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
bitwiseFuncMap := map[string]func(a, b int) int{
|
|
"BITAND": func(a, b int) int { return a & b },
|
|
"BITLSHIFT": func(a, b int) int { return a << uint(b) },
|
|
"BITOR": func(a, b int) int { return a | b },
|
|
"BITRSHIFT": func(a, b int) int { return a >> uint(b) },
|
|
"BITXOR": func(a, b int) int { return a ^ b },
|
|
}
|
|
bitwiseFunc := bitwiseFuncMap[name]
|
|
return newNumberFormulaArg(float64(bitwiseFunc(int(num1.Number), int(num2.Number))))
|
|
}
|
|
|
|
// COMPLEX function takes two arguments, representing the real and the
|
|
// imaginary coefficients of a complex number, and from these, creates a
|
|
// complex number. The syntax of the function is:
|
|
//
|
|
// COMPLEX(real_num,i_num,[suffix])
|
|
func (fn *formulaFuncs) COMPLEX(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMPLEX requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMPLEX allows at most 3 arguments")
|
|
}
|
|
realNum, i, suffix := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Front().Next().Value.(formulaArg).ToNumber(), "i"
|
|
if realNum.Type != ArgNumber {
|
|
return realNum
|
|
}
|
|
if i.Type != ArgNumber {
|
|
return i
|
|
}
|
|
if argsList.Len() == 3 {
|
|
if suffix = strings.ToLower(argsList.Back().Value.(formulaArg).Value()); suffix != "i" && suffix != "j" {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
return newStringFormulaArg(cmplx2str(complex(realNum.Number, i.Number), suffix))
|
|
}
|
|
|
|
// cmplx2str replace complex number string characters.
|
|
func cmplx2str(num complex128, suffix string) string {
|
|
realPart, imagPart := fmt.Sprint(real(num)), fmt.Sprint(imag(num))
|
|
isNum, i, decimal := isNumeric(realPart)
|
|
if isNum && i > 15 {
|
|
realPart = strconv.FormatFloat(decimal, 'G', 15, 64)
|
|
}
|
|
isNum, i, decimal = isNumeric(imagPart)
|
|
if isNum && i > 15 {
|
|
imagPart = strconv.FormatFloat(decimal, 'G', 15, 64)
|
|
}
|
|
c := realPart
|
|
if imag(num) > 0 {
|
|
c += "+"
|
|
}
|
|
if imag(num) != 0 {
|
|
c += imagPart + "i"
|
|
}
|
|
c = strings.TrimPrefix(c, "(")
|
|
c = strings.TrimPrefix(c, "+0+")
|
|
c = strings.TrimPrefix(c, "-0+")
|
|
c = strings.TrimSuffix(c, ")")
|
|
c = strings.TrimPrefix(c, "0+")
|
|
if strings.HasPrefix(c, "0-") {
|
|
c = "-" + strings.TrimPrefix(c, "0-")
|
|
}
|
|
c = strings.TrimPrefix(c, "0+")
|
|
c = strings.TrimSuffix(c, "+0i")
|
|
c = strings.TrimSuffix(c, "-0i")
|
|
c = strings.NewReplacer("+1i", "+i", "-1i", "-i").Replace(c)
|
|
c = strings.ReplaceAll(c, "i", suffix)
|
|
return c
|
|
}
|
|
|
|
// str2cmplx convert complex number string characters.
|
|
func str2cmplx(c string) string {
|
|
c = strings.ReplaceAll(c, "j", "i")
|
|
if c == "i" {
|
|
c = "1i"
|
|
}
|
|
c = strings.NewReplacer("+i", "+1i", "-i", "-1i").Replace(c)
|
|
return c
|
|
}
|
|
|
|
// conversionUnit defined unit info for conversion.
|
|
type conversionUnit struct {
|
|
group uint8
|
|
allowPrefix bool
|
|
}
|
|
|
|
// conversionUnits maps info list for unit conversion, that can be used in
|
|
// formula function CONVERT.
|
|
var conversionUnits = map[string]conversionUnit{
|
|
// weight and mass
|
|
"g": {group: categoryWeightAndMass, allowPrefix: true},
|
|
"sg": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"lbm": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"u": {group: categoryWeightAndMass, allowPrefix: true},
|
|
"ozm": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"grain": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"cwt": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"shweight": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"uk_cwt": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"lcwt": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"hweight": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"stone": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"ton": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"uk_ton": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"LTON": {group: categoryWeightAndMass, allowPrefix: false},
|
|
"brton": {group: categoryWeightAndMass, allowPrefix: false},
|
|
// distance
|
|
"m": {group: categoryDistance, allowPrefix: true},
|
|
"mi": {group: categoryDistance, allowPrefix: false},
|
|
"Nmi": {group: categoryDistance, allowPrefix: false},
|
|
"in": {group: categoryDistance, allowPrefix: false},
|
|
"ft": {group: categoryDistance, allowPrefix: false},
|
|
"yd": {group: categoryDistance, allowPrefix: false},
|
|
"ang": {group: categoryDistance, allowPrefix: true},
|
|
"ell": {group: categoryDistance, allowPrefix: false},
|
|
"ly": {group: categoryDistance, allowPrefix: false},
|
|
"parsec": {group: categoryDistance, allowPrefix: false},
|
|
"pc": {group: categoryDistance, allowPrefix: false},
|
|
"Pica": {group: categoryDistance, allowPrefix: false},
|
|
"Picapt": {group: categoryDistance, allowPrefix: false},
|
|
"pica": {group: categoryDistance, allowPrefix: false},
|
|
"survey_mi": {group: categoryDistance, allowPrefix: false},
|
|
// time
|
|
"yr": {group: categoryTime, allowPrefix: false},
|
|
"day": {group: categoryTime, allowPrefix: false},
|
|
"d": {group: categoryTime, allowPrefix: false},
|
|
"hr": {group: categoryTime, allowPrefix: false},
|
|
"mn": {group: categoryTime, allowPrefix: false},
|
|
"min": {group: categoryTime, allowPrefix: false},
|
|
"sec": {group: categoryTime, allowPrefix: true},
|
|
"s": {group: categoryTime, allowPrefix: true},
|
|
// pressure
|
|
"Pa": {group: categoryPressure, allowPrefix: true},
|
|
"p": {group: categoryPressure, allowPrefix: true},
|
|
"atm": {group: categoryPressure, allowPrefix: true},
|
|
"at": {group: categoryPressure, allowPrefix: true},
|
|
"mmHg": {group: categoryPressure, allowPrefix: true},
|
|
"psi": {group: categoryPressure, allowPrefix: true},
|
|
"Torr": {group: categoryPressure, allowPrefix: true},
|
|
// force
|
|
"N": {group: categoryForce, allowPrefix: true},
|
|
"dyn": {group: categoryForce, allowPrefix: true},
|
|
"dy": {group: categoryForce, allowPrefix: true},
|
|
"lbf": {group: categoryForce, allowPrefix: false},
|
|
"pond": {group: categoryForce, allowPrefix: true},
|
|
// energy
|
|
"J": {group: categoryEnergy, allowPrefix: true},
|
|
"e": {group: categoryEnergy, allowPrefix: true},
|
|
"c": {group: categoryEnergy, allowPrefix: true},
|
|
"cal": {group: categoryEnergy, allowPrefix: true},
|
|
"eV": {group: categoryEnergy, allowPrefix: true},
|
|
"ev": {group: categoryEnergy, allowPrefix: true},
|
|
"HPh": {group: categoryEnergy, allowPrefix: false},
|
|
"hh": {group: categoryEnergy, allowPrefix: false},
|
|
"Wh": {group: categoryEnergy, allowPrefix: true},
|
|
"wh": {group: categoryEnergy, allowPrefix: true},
|
|
"flb": {group: categoryEnergy, allowPrefix: false},
|
|
"BTU": {group: categoryEnergy, allowPrefix: false},
|
|
"btu": {group: categoryEnergy, allowPrefix: false},
|
|
// power
|
|
"HP": {group: categoryPower, allowPrefix: false},
|
|
"h": {group: categoryPower, allowPrefix: false},
|
|
"W": {group: categoryPower, allowPrefix: true},
|
|
"w": {group: categoryPower, allowPrefix: true},
|
|
"PS": {group: categoryPower, allowPrefix: false},
|
|
"T": {group: categoryMagnetism, allowPrefix: true},
|
|
"ga": {group: categoryMagnetism, allowPrefix: true},
|
|
// temperature
|
|
"C": {group: categoryTemperature, allowPrefix: false},
|
|
"cel": {group: categoryTemperature, allowPrefix: false},
|
|
"F": {group: categoryTemperature, allowPrefix: false},
|
|
"fah": {group: categoryTemperature, allowPrefix: false},
|
|
"K": {group: categoryTemperature, allowPrefix: false},
|
|
"kel": {group: categoryTemperature, allowPrefix: false},
|
|
"Rank": {group: categoryTemperature, allowPrefix: false},
|
|
"Reau": {group: categoryTemperature, allowPrefix: false},
|
|
// volume
|
|
"l": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"L": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"lt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"tsp": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"tspm": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"tbs": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"oz": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"cup": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"pt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"us_pt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"uk_pt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"qt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"uk_qt": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"gal": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"uk_gal": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"ang3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"ang^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"barrel": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"bushel": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"in3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"in^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"ft3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"ft^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"ly3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"ly^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"m3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"m^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: true},
|
|
"mi3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"mi^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"yd3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"yd^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Nmi3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Nmi^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Pica3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Pica^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Picapt3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"Picapt^3": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"GRT": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"regton": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
"MTON": {group: categoryVolumeAndLiquidMeasure, allowPrefix: false},
|
|
// area
|
|
"ha": {group: categoryArea, allowPrefix: true},
|
|
"uk_acre": {group: categoryArea, allowPrefix: false},
|
|
"us_acre": {group: categoryArea, allowPrefix: false},
|
|
"ang2": {group: categoryArea, allowPrefix: true},
|
|
"ang^2": {group: categoryArea, allowPrefix: true},
|
|
"ar": {group: categoryArea, allowPrefix: true},
|
|
"ft2": {group: categoryArea, allowPrefix: false},
|
|
"ft^2": {group: categoryArea, allowPrefix: false},
|
|
"in2": {group: categoryArea, allowPrefix: false},
|
|
"in^2": {group: categoryArea, allowPrefix: false},
|
|
"ly2": {group: categoryArea, allowPrefix: false},
|
|
"ly^2": {group: categoryArea, allowPrefix: false},
|
|
"m2": {group: categoryArea, allowPrefix: true},
|
|
"m^2": {group: categoryArea, allowPrefix: true},
|
|
"Morgen": {group: categoryArea, allowPrefix: false},
|
|
"mi2": {group: categoryArea, allowPrefix: false},
|
|
"mi^2": {group: categoryArea, allowPrefix: false},
|
|
"Nmi2": {group: categoryArea, allowPrefix: false},
|
|
"Nmi^2": {group: categoryArea, allowPrefix: false},
|
|
"Pica2": {group: categoryArea, allowPrefix: false},
|
|
"Pica^2": {group: categoryArea, allowPrefix: false},
|
|
"Picapt2": {group: categoryArea, allowPrefix: false},
|
|
"Picapt^2": {group: categoryArea, allowPrefix: false},
|
|
"yd2": {group: categoryArea, allowPrefix: false},
|
|
"yd^2": {group: categoryArea, allowPrefix: false},
|
|
// information
|
|
"byte": {group: categoryInformation, allowPrefix: true},
|
|
"bit": {group: categoryInformation, allowPrefix: true},
|
|
// speed
|
|
"m/s": {group: categorySpeed, allowPrefix: true},
|
|
"m/sec": {group: categorySpeed, allowPrefix: true},
|
|
"m/h": {group: categorySpeed, allowPrefix: true},
|
|
"m/hr": {group: categorySpeed, allowPrefix: true},
|
|
"mph": {group: categorySpeed, allowPrefix: false},
|
|
"admkn": {group: categorySpeed, allowPrefix: false},
|
|
"kn": {group: categorySpeed, allowPrefix: false},
|
|
}
|
|
|
|
// unitConversions maps details of the Units of measure conversion factors,
|
|
// organised by group.
|
|
var unitConversions = map[byte]map[string]float64{
|
|
// conversion uses gram (g) as an intermediate unit
|
|
categoryWeightAndMass: {
|
|
"g": 1,
|
|
"sg": 6.85217658567918e-05,
|
|
"lbm": 2.20462262184878e-03,
|
|
"u": 6.02214179421676e+23,
|
|
"ozm": 3.52739619495804e-02,
|
|
"grain": 1.54323583529414e+01,
|
|
"cwt": 2.20462262184878e-05,
|
|
"shweight": 2.20462262184878e-05,
|
|
"uk_cwt": 1.96841305522212e-05,
|
|
"lcwt": 1.96841305522212e-05,
|
|
"hweight": 1.96841305522212e-05,
|
|
"stone": 1.57473044417770e-04,
|
|
"ton": 1.10231131092439e-06,
|
|
"uk_ton": 9.84206527611061e-07,
|
|
"LTON": 9.84206527611061e-07,
|
|
"brton": 9.84206527611061e-07,
|
|
},
|
|
// conversion uses meter (m) as an intermediate unit
|
|
categoryDistance: {
|
|
"m": 1,
|
|
"mi": 6.21371192237334e-04,
|
|
"Nmi": 5.39956803455724e-04,
|
|
"in": 3.93700787401575e+01,
|
|
"ft": 3.28083989501312e+00,
|
|
"yd": 1.09361329833771e+00,
|
|
"ang": 1.0e+10,
|
|
"ell": 8.74890638670166e-01,
|
|
"ly": 1.05700083402462e-16,
|
|
"parsec": 3.24077928966473e-17,
|
|
"pc": 3.24077928966473e-17,
|
|
"Pica": 2.83464566929134e+03,
|
|
"Picapt": 2.83464566929134e+03,
|
|
"pica": 2.36220472440945e+02,
|
|
"survey_mi": 6.21369949494950e-04,
|
|
},
|
|
// conversion uses second (s) as an intermediate unit
|
|
categoryTime: {
|
|
"yr": 3.16880878140289e-08,
|
|
"day": 1.15740740740741e-05,
|
|
"d": 1.15740740740741e-05,
|
|
"hr": 2.77777777777778e-04,
|
|
"mn": 1.66666666666667e-02,
|
|
"min": 1.66666666666667e-02,
|
|
"sec": 1,
|
|
"s": 1,
|
|
},
|
|
// conversion uses Pascal (Pa) as an intermediate unit
|
|
categoryPressure: {
|
|
"Pa": 1,
|
|
"p": 1,
|
|
"atm": 9.86923266716013e-06,
|
|
"at": 9.86923266716013e-06,
|
|
"mmHg": 7.50063755419211e-03,
|
|
"psi": 1.45037737730209e-04,
|
|
"Torr": 7.50061682704170e-03,
|
|
},
|
|
// conversion uses Newton (N) as an intermediate unit
|
|
categoryForce: {
|
|
"N": 1,
|
|
"dyn": 1.0e+5,
|
|
"dy": 1.0e+5,
|
|
"lbf": 2.24808923655339e-01,
|
|
"pond": 1.01971621297793e+02,
|
|
},
|
|
// conversion uses Joule (J) as an intermediate unit
|
|
categoryEnergy: {
|
|
"J": 1,
|
|
"e": 9.99999519343231e+06,
|
|
"c": 2.39006249473467e-01,
|
|
"cal": 2.38846190642017e-01,
|
|
"eV": 6.24145700000000e+18,
|
|
"ev": 6.24145700000000e+18,
|
|
"HPh": 3.72506430801000e-07,
|
|
"hh": 3.72506430801000e-07,
|
|
"Wh": 2.77777916238711e-04,
|
|
"wh": 2.77777916238711e-04,
|
|
"flb": 2.37304222192651e+01,
|
|
"BTU": 9.47815067349015e-04,
|
|
"btu": 9.47815067349015e-04,
|
|
},
|
|
// conversion uses Horsepower (HP) as an intermediate unit
|
|
categoryPower: {
|
|
"HP": 1,
|
|
"h": 1,
|
|
"W": 7.45699871582270e+02,
|
|
"w": 7.45699871582270e+02,
|
|
"PS": 1.01386966542400e+00,
|
|
},
|
|
// conversion uses Tesla (T) as an intermediate unit
|
|
categoryMagnetism: {
|
|
"T": 1,
|
|
"ga": 10000,
|
|
},
|
|
// conversion uses litre (l) as an intermediate unit
|
|
categoryVolumeAndLiquidMeasure: {
|
|
"l": 1,
|
|
"L": 1,
|
|
"lt": 1,
|
|
"tsp": 2.02884136211058e+02,
|
|
"tspm": 2.0e+02,
|
|
"tbs": 6.76280454036860e+01,
|
|
"oz": 3.38140227018430e+01,
|
|
"cup": 4.22675283773038e+00,
|
|
"pt": 2.11337641886519e+00,
|
|
"us_pt": 2.11337641886519e+00,
|
|
"uk_pt": 1.75975398639270e+00,
|
|
"qt": 1.05668820943259e+00,
|
|
"uk_qt": 8.79876993196351e-01,
|
|
"gal": 2.64172052358148e-01,
|
|
"uk_gal": 2.19969248299088e-01,
|
|
"ang3": 1.0e+27,
|
|
"ang^3": 1.0e+27,
|
|
"barrel": 6.28981077043211e-03,
|
|
"bushel": 2.83775932584017e-02,
|
|
"in3": 6.10237440947323e+01,
|
|
"in^3": 6.10237440947323e+01,
|
|
"ft3": 3.53146667214886e-02,
|
|
"ft^3": 3.53146667214886e-02,
|
|
"ly3": 1.18093498844171e-51,
|
|
"ly^3": 1.18093498844171e-51,
|
|
"m3": 1.0e-03,
|
|
"m^3": 1.0e-03,
|
|
"mi3": 2.39912758578928e-13,
|
|
"mi^3": 2.39912758578928e-13,
|
|
"yd3": 1.30795061931439e-03,
|
|
"yd^3": 1.30795061931439e-03,
|
|
"Nmi3": 1.57426214685811e-13,
|
|
"Nmi^3": 1.57426214685811e-13,
|
|
"Pica3": 2.27769904358706e+07,
|
|
"Pica^3": 2.27769904358706e+07,
|
|
"Picapt3": 2.27769904358706e+07,
|
|
"Picapt^3": 2.27769904358706e+07,
|
|
"GRT": 3.53146667214886e-04,
|
|
"regton": 3.53146667214886e-04,
|
|
"MTON": 8.82866668037215e-04,
|
|
},
|
|
// conversion uses hectare (ha) as an intermediate unit
|
|
categoryArea: {
|
|
"ha": 1,
|
|
"uk_acre": 2.47105381467165e+00,
|
|
"us_acre": 2.47104393046628e+00,
|
|
"ang2": 1.0e+24,
|
|
"ang^2": 1.0e+24,
|
|
"ar": 1.0e+02,
|
|
"ft2": 1.07639104167097e+05,
|
|
"ft^2": 1.07639104167097e+05,
|
|
"in2": 1.55000310000620e+07,
|
|
"in^2": 1.55000310000620e+07,
|
|
"ly2": 1.11725076312873e-28,
|
|
"ly^2": 1.11725076312873e-28,
|
|
"m2": 1.0e+04,
|
|
"m^2": 1.0e+04,
|
|
"Morgen": 4.0e+00,
|
|
"mi2": 3.86102158542446e-03,
|
|
"mi^2": 3.86102158542446e-03,
|
|
"Nmi2": 2.91553349598123e-03,
|
|
"Nmi^2": 2.91553349598123e-03,
|
|
"Pica2": 8.03521607043214e+10,
|
|
"Pica^2": 8.03521607043214e+10,
|
|
"Picapt2": 8.03521607043214e+10,
|
|
"Picapt^2": 8.03521607043214e+10,
|
|
"yd2": 1.19599004630108e+04,
|
|
"yd^2": 1.19599004630108e+04,
|
|
},
|
|
// conversion uses bit (bit) as an intermediate unit
|
|
categoryInformation: {
|
|
"bit": 1,
|
|
"byte": 0.125,
|
|
},
|
|
// conversion uses Meters per Second (m/s) as an intermediate unit
|
|
categorySpeed: {
|
|
"m/s": 1,
|
|
"m/sec": 1,
|
|
"m/h": 3.60e+03,
|
|
"m/hr": 3.60e+03,
|
|
"mph": 2.23693629205440e+00,
|
|
"admkn": 1.94260256941567e+00,
|
|
"kn": 1.94384449244060e+00,
|
|
},
|
|
}
|
|
|
|
// conversionMultipliers maps details of the Multiplier prefixes that can be
|
|
// used with Units of Measure in CONVERT.
|
|
var conversionMultipliers = map[string]float64{
|
|
"Y": 1e24,
|
|
"Z": 1e21,
|
|
"E": 1e18,
|
|
"P": 1e15,
|
|
"T": 1e12,
|
|
"G": 1e9,
|
|
"M": 1e6,
|
|
"k": 1e3,
|
|
"h": 1e2,
|
|
"e": 1e1,
|
|
"da": 1e1,
|
|
"d": 1e-1,
|
|
"c": 1e-2,
|
|
"m": 1e-3,
|
|
"u": 1e-6,
|
|
"n": 1e-9,
|
|
"p": 1e-12,
|
|
"f": 1e-15,
|
|
"a": 1e-18,
|
|
"z": 1e-21,
|
|
"y": 1e-24,
|
|
"Yi": math.Pow(2, 80),
|
|
"Zi": math.Pow(2, 70),
|
|
"Ei": math.Pow(2, 60),
|
|
"Pi": math.Pow(2, 50),
|
|
"Ti": math.Pow(2, 40),
|
|
"Gi": math.Pow(2, 30),
|
|
"Mi": math.Pow(2, 20),
|
|
"ki": math.Pow(2, 10),
|
|
}
|
|
|
|
// getUnitDetails check and returns the unit of measure details.
|
|
func getUnitDetails(uom string) (unit string, catgory byte, res float64, ok bool) {
|
|
if len(uom) == 0 {
|
|
ok = false
|
|
return
|
|
}
|
|
if unit, ok := conversionUnits[uom]; ok {
|
|
return uom, unit.group, 1, ok
|
|
}
|
|
// 1 character standard metric multiplier prefixes
|
|
multiplierType := uom[:1]
|
|
uom = uom[1:]
|
|
conversionUnit, ok1 := conversionUnits[uom]
|
|
multiplier, ok2 := conversionMultipliers[multiplierType]
|
|
if ok1 && ok2 {
|
|
if !conversionUnit.allowPrefix {
|
|
ok = false
|
|
return
|
|
}
|
|
unitCategory := conversionUnit.group
|
|
return uom, unitCategory, multiplier, true
|
|
}
|
|
// 2 character standard and binary metric multiplier prefixes
|
|
if len(uom) > 0 {
|
|
multiplierType += uom[:1]
|
|
uom = uom[1:]
|
|
}
|
|
conversionUnit, ok1 = conversionUnits[uom]
|
|
multiplier, ok2 = conversionMultipliers[multiplierType]
|
|
if ok1 && ok2 {
|
|
if !conversionUnit.allowPrefix {
|
|
ok = false
|
|
return
|
|
}
|
|
unitCategory := conversionUnit.group
|
|
return uom, unitCategory, multiplier, true
|
|
}
|
|
ok = false
|
|
return
|
|
}
|
|
|
|
// resolveTemperatureSynonyms returns unit of measure according to a given
|
|
// temperature synonyms.
|
|
func resolveTemperatureSynonyms(uom string) string {
|
|
switch uom {
|
|
case "fah":
|
|
return "F"
|
|
case "cel":
|
|
return "C"
|
|
case "kel":
|
|
return "K"
|
|
}
|
|
return uom
|
|
}
|
|
|
|
// convertTemperature returns converted temperature by a given unit of measure.
|
|
func convertTemperature(fromUOM, toUOM string, value float64) float64 {
|
|
fromUOM = resolveTemperatureSynonyms(fromUOM)
|
|
toUOM = resolveTemperatureSynonyms(toUOM)
|
|
if fromUOM == toUOM {
|
|
return value
|
|
}
|
|
// convert to Kelvin
|
|
switch fromUOM {
|
|
case "F":
|
|
value = (value-32)/1.8 + 273.15
|
|
case "C":
|
|
value += 273.15
|
|
case "Rank":
|
|
value /= 1.8
|
|
case "Reau":
|
|
value = value*1.25 + 273.15
|
|
}
|
|
// convert from Kelvin
|
|
switch toUOM {
|
|
case "F":
|
|
value = (value-273.15)*1.8 + 32
|
|
case "C":
|
|
value -= 273.15
|
|
case "Rank":
|
|
value *= 1.8
|
|
case "Reau":
|
|
value = (value - 273.15) * 0.8
|
|
}
|
|
return value
|
|
}
|
|
|
|
// CONVERT function converts a number from one unit type (e.g. Yards) to
|
|
// another unit type (e.g. Meters). The syntax of the function is:
|
|
//
|
|
// CONVERT(number,from_unit,to_unit)
|
|
func (fn *formulaFuncs) CONVERT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CONVERT requires 3 arguments")
|
|
}
|
|
num := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return num
|
|
}
|
|
fromUOM, fromCategory, fromMultiplier, ok1 := getUnitDetails(argsList.Front().Next().Value.(formulaArg).Value())
|
|
toUOM, toCategory, toMultiplier, ok2 := getUnitDetails(argsList.Back().Value.(formulaArg).Value())
|
|
if !ok1 || !ok2 || fromCategory != toCategory {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
val := num.Number * fromMultiplier
|
|
if fromUOM == toUOM && fromMultiplier == toMultiplier {
|
|
return newNumberFormulaArg(val / fromMultiplier)
|
|
} else if fromUOM == toUOM {
|
|
return newNumberFormulaArg(val / toMultiplier)
|
|
} else if fromCategory == categoryTemperature {
|
|
return newNumberFormulaArg(convertTemperature(fromUOM, toUOM, val))
|
|
}
|
|
fromConversion := unitConversions[fromCategory][fromUOM]
|
|
toConversion := unitConversions[fromCategory][toUOM]
|
|
baseValue := val * (1 / fromConversion)
|
|
return newNumberFormulaArg((baseValue * toConversion) / toMultiplier)
|
|
}
|
|
|
|
// DEC2BIN function converts a decimal number into a Binary (Base 2) number.
|
|
// The syntax of the function is:
|
|
//
|
|
// DEC2BIN(number,[places])
|
|
func (fn *formulaFuncs) DEC2BIN(argsList *list.List) formulaArg {
|
|
return fn.dec2x("DEC2BIN", argsList)
|
|
}
|
|
|
|
// DEC2HEX function converts a decimal number into a Hexadecimal (Base 16)
|
|
// number. The syntax of the function is:
|
|
//
|
|
// DEC2HEX(number,[places])
|
|
func (fn *formulaFuncs) DEC2HEX(argsList *list.List) formulaArg {
|
|
return fn.dec2x("DEC2HEX", argsList)
|
|
}
|
|
|
|
// DEC2OCT function converts a decimal number into an Octal (Base 8) number.
|
|
// The syntax of the function is:
|
|
//
|
|
// DEC2OCT(number,[places])
|
|
func (fn *formulaFuncs) DEC2OCT(argsList *list.List) formulaArg {
|
|
return fn.dec2x("DEC2OCT", argsList)
|
|
}
|
|
|
|
// dec2x is an implementation of the formula functions DEC2BIN, DEC2HEX and
|
|
// DEC2OCT.
|
|
func (fn *formulaFuncs) dec2x(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 2 arguments", name))
|
|
}
|
|
decimal := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if decimal.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, decimal.Error)
|
|
}
|
|
maxLimitMap := map[string]float64{
|
|
"DEC2BIN": 511,
|
|
"HEX2BIN": 511,
|
|
"OCT2BIN": 511,
|
|
"BIN2HEX": 549755813887,
|
|
"DEC2HEX": 549755813887,
|
|
"OCT2HEX": 549755813887,
|
|
"BIN2OCT": 536870911,
|
|
"DEC2OCT": 536870911,
|
|
"HEX2OCT": 536870911,
|
|
}
|
|
minLimitMap := map[string]float64{
|
|
"DEC2BIN": -512,
|
|
"HEX2BIN": -512,
|
|
"OCT2BIN": -512,
|
|
"BIN2HEX": -549755813888,
|
|
"DEC2HEX": -549755813888,
|
|
"OCT2HEX": -549755813888,
|
|
"BIN2OCT": -536870912,
|
|
"DEC2OCT": -536870912,
|
|
"HEX2OCT": -536870912,
|
|
}
|
|
baseMap := map[string]int{
|
|
"DEC2BIN": 2,
|
|
"HEX2BIN": 2,
|
|
"OCT2BIN": 2,
|
|
"BIN2HEX": 16,
|
|
"DEC2HEX": 16,
|
|
"OCT2HEX": 16,
|
|
"BIN2OCT": 8,
|
|
"DEC2OCT": 8,
|
|
"HEX2OCT": 8,
|
|
}
|
|
maxLimit, minLimit := maxLimitMap[name], minLimitMap[name]
|
|
base := baseMap[name]
|
|
if decimal.Number < minLimit || decimal.Number > maxLimit {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
n := int64(decimal.Number)
|
|
binary := strconv.FormatUint(*(*uint64)(unsafe.Pointer(&n)), base)
|
|
if argsList.Len() == 2 {
|
|
places := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if places.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, places.Error)
|
|
}
|
|
binaryPlaces := len(binary)
|
|
if places.Number < 0 || places.Number > 10 || binaryPlaces > int(places.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(fmt.Sprintf("%s%s", strings.Repeat("0", int(places.Number)-binaryPlaces), binary)))
|
|
}
|
|
if decimal.Number < 0 && len(binary) > 10 {
|
|
return newStringFormulaArg(strings.ToUpper(binary[len(binary)-10:]))
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(binary))
|
|
}
|
|
|
|
// DELTA function tests two numbers for equality and returns the Kronecker
|
|
// Delta. i.e. the function returns 1 if the two supplied numbers are equal
|
|
// and 0 otherwise. The syntax of the function is:
|
|
//
|
|
// DELTA(number1,[number2])
|
|
func (fn *formulaFuncs) DELTA(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DELTA requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DELTA allows at most 2 arguments")
|
|
}
|
|
number1 := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number1.Type != ArgNumber {
|
|
return number1
|
|
}
|
|
number2 := newNumberFormulaArg(0)
|
|
if argsList.Len() == 2 {
|
|
if number2 = argsList.Back().Value.(formulaArg).ToNumber(); number2.Type != ArgNumber {
|
|
return number2
|
|
}
|
|
}
|
|
return newBoolFormulaArg(number1.Number == number2.Number).ToNumber()
|
|
}
|
|
|
|
// ERF function calculates the Error Function, integrated between two supplied
|
|
// limits. The syntax of the function is:
|
|
//
|
|
// ERF(lower_limit,[upper_limit])
|
|
func (fn *formulaFuncs) ERF(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ERF requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ERF allows at most 2 arguments")
|
|
}
|
|
lower := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if lower.Type != ArgNumber {
|
|
return lower
|
|
}
|
|
if argsList.Len() == 2 {
|
|
upper := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if upper.Type != ArgNumber {
|
|
return upper
|
|
}
|
|
return newNumberFormulaArg(math.Erf(upper.Number) - math.Erf(lower.Number))
|
|
}
|
|
return newNumberFormulaArg(math.Erf(lower.Number))
|
|
}
|
|
|
|
// ERFdotPRECISE function calculates the Error Function, integrated between a
|
|
// supplied lower or upper limit and 0. The syntax of the function is:
|
|
//
|
|
// ERF.PRECISE(x)
|
|
func (fn *formulaFuncs) ERFdotPRECISE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ERF.PRECISE requires 1 argument")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
return newNumberFormulaArg(math.Erf(x.Number))
|
|
}
|
|
|
|
// erfc is an implementation of the formula functions ERFC and ERFC.PRECISE.
|
|
func (fn *formulaFuncs) erfc(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 1 argument", name))
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
return newNumberFormulaArg(math.Erfc(x.Number))
|
|
}
|
|
|
|
// ERFC function calculates the Complementary Error Function, integrated
|
|
// between a supplied lower limit and infinity. The syntax of the function
|
|
// is:
|
|
//
|
|
// ERFC(x)
|
|
func (fn *formulaFuncs) ERFC(argsList *list.List) formulaArg {
|
|
return fn.erfc("ERFC", argsList)
|
|
}
|
|
|
|
// ERFCdotPRECISE function calculates the Complementary Error Function,
|
|
// integrated between a supplied lower limit and infinity. The syntax of the
|
|
// function is:
|
|
//
|
|
// ERFC(x)
|
|
func (fn *formulaFuncs) ERFCdotPRECISE(argsList *list.List) formulaArg {
|
|
return fn.erfc("ERFC.PRECISE", argsList)
|
|
}
|
|
|
|
// GESTEP unction tests whether a supplied number is greater than a supplied
|
|
// step size and returns. The syntax of the function is:
|
|
//
|
|
// GESTEP(number,[step])
|
|
func (fn *formulaFuncs) GESTEP(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GESTEP requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GESTEP allows at most 2 arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return number
|
|
}
|
|
step := newNumberFormulaArg(0)
|
|
if argsList.Len() == 2 {
|
|
if step = argsList.Back().Value.(formulaArg).ToNumber(); step.Type != ArgNumber {
|
|
return step
|
|
}
|
|
}
|
|
return newBoolFormulaArg(number.Number >= step.Number).ToNumber()
|
|
}
|
|
|
|
// HEX2BIN function converts a Hexadecimal (Base 16) number into a Binary
|
|
// (Base 2) number. The syntax of the function is:
|
|
//
|
|
// HEX2BIN(number,[places])
|
|
func (fn *formulaFuncs) HEX2BIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HEX2BIN requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HEX2BIN allows at most 2 arguments")
|
|
}
|
|
decimal, newList := fn.hex2dec(argsList.Front().Value.(formulaArg).Value()), list.New()
|
|
if decimal.Type != ArgNumber {
|
|
return decimal
|
|
}
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("HEX2BIN", newList)
|
|
}
|
|
|
|
// HEX2DEC function converts a hexadecimal (a base-16 number) into a decimal
|
|
// number. The syntax of the function is:
|
|
//
|
|
// HEX2DEC(number)
|
|
func (fn *formulaFuncs) HEX2DEC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HEX2DEC requires 1 numeric argument")
|
|
}
|
|
return fn.hex2dec(argsList.Front().Value.(formulaArg).Value())
|
|
}
|
|
|
|
// HEX2OCT function converts a Hexadecimal (Base 16) number into an Octal
|
|
// (Base 8) number. The syntax of the function is:
|
|
//
|
|
// HEX2OCT(number,[places])
|
|
func (fn *formulaFuncs) HEX2OCT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HEX2OCT requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HEX2OCT allows at most 2 arguments")
|
|
}
|
|
decimal, newList := fn.hex2dec(argsList.Front().Value.(formulaArg).Value()), list.New()
|
|
if decimal.Type != ArgNumber {
|
|
return decimal
|
|
}
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("HEX2OCT", newList)
|
|
}
|
|
|
|
// hex2dec is an implementation of the formula function HEX2DEC.
|
|
func (fn *formulaFuncs) hex2dec(number string) formulaArg {
|
|
decimal, length := 0.0, len(number)
|
|
for i := length; i > 0; i-- {
|
|
num, err := strconv.ParseInt(string(number[length-i]), 16, 64)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
if i == 10 && string(number[length-i]) == "F" {
|
|
decimal += math.Pow(-16.0, float64(i-1))
|
|
continue
|
|
}
|
|
decimal += float64(num) * math.Pow(16.0, float64(i-1))
|
|
}
|
|
return newNumberFormulaArg(decimal)
|
|
}
|
|
|
|
// IMABS function returns the absolute value (the modulus) of a complex
|
|
// number. The syntax of the function is:
|
|
//
|
|
// IMABS(inumber)
|
|
func (fn *formulaFuncs) IMABS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMABS requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newNumberFormulaArg(cmplx.Abs(inumber))
|
|
}
|
|
|
|
// IMAGINARY function returns the imaginary coefficient of a supplied complex
|
|
// number. The syntax of the function is:
|
|
//
|
|
// IMAGINARY(inumber)
|
|
func (fn *formulaFuncs) IMAGINARY(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMAGINARY requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newNumberFormulaArg(imag(inumber))
|
|
}
|
|
|
|
// IMARGUMENT function returns the phase (also called the argument) of a
|
|
// supplied complex number. The syntax of the function is:
|
|
//
|
|
// IMARGUMENT(inumber)
|
|
func (fn *formulaFuncs) IMARGUMENT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMARGUMENT requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newNumberFormulaArg(cmplx.Phase(inumber))
|
|
}
|
|
|
|
// IMCONJUGATE function returns the complex conjugate of a supplied complex
|
|
// number. The syntax of the function is:
|
|
//
|
|
// IMCONJUGATE(inumber)
|
|
func (fn *formulaFuncs) IMCONJUGATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCONJUGATE requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Conj(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMCOS function returns the cosine of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMCOS(inumber)
|
|
func (fn *formulaFuncs) IMCOS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCOS requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Cos(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMCOSH function returns the hyperbolic cosine of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMCOSH(inumber)
|
|
func (fn *formulaFuncs) IMCOSH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCOSH requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Cosh(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMCOT function returns the cotangent of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMCOT(inumber)
|
|
func (fn *formulaFuncs) IMCOT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCOT requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Cot(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMCSC function returns the cosecant of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMCSC(inumber)
|
|
func (fn *formulaFuncs) IMCSC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCSC requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := 1 / cmplx.Sin(inumber)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMCSCH function returns the hyperbolic cosecant of a supplied complex
|
|
// number. The syntax of the function is:
|
|
//
|
|
// IMCSCH(inumber)
|
|
func (fn *formulaFuncs) IMCSCH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMCSCH requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := 1 / cmplx.Sinh(inumber)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMDIV function calculates the quotient of two complex numbers (i.e. divides
|
|
// one complex number by another). The syntax of the function is:
|
|
//
|
|
// IMDIV(inumber1,inumber2)
|
|
func (fn *formulaFuncs) IMDIV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMDIV requires 2 arguments")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber1, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
inumber2, err := strconv.ParseComplex(str2cmplx(argsList.Back().Value.(formulaArg).Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := inumber1 / inumber2
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMEXP function returns the exponential of a supplied complex number. The
|
|
// syntax of the function is:
|
|
//
|
|
// IMEXP(inumber)
|
|
func (fn *formulaFuncs) IMEXP(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMEXP requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Exp(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMLN function returns the natural logarithm of a supplied complex number.
|
|
// The syntax of the function is:
|
|
//
|
|
// IMLN(inumber)
|
|
func (fn *formulaFuncs) IMLN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMLN requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := cmplx.Log(inumber)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMLOG10 function returns the common (base 10) logarithm of a supplied
|
|
// complex number. The syntax of the function is:
|
|
//
|
|
// IMLOG10(inumber)
|
|
func (fn *formulaFuncs) IMLOG10(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMLOG10 requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := cmplx.Log10(inumber)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMLOG2 function calculates the base 2 logarithm of a supplied complex
|
|
// number. The syntax of the function is:
|
|
//
|
|
// IMLOG2(inumber)
|
|
func (fn *formulaFuncs) IMLOG2(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMLOG2 requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
num := cmplx.Log(inumber)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num/cmplx.Log(2), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMPOWER function returns a supplied complex number, raised to a given
|
|
// power. The syntax of the function is:
|
|
//
|
|
// IMPOWER(inumber,number)
|
|
func (fn *formulaFuncs) IMPOWER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMPOWER requires 2 arguments")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
number, err := strconv.ParseComplex(str2cmplx(argsList.Back().Value.(formulaArg).Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
if inumber == 0 && number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
num := cmplx.Pow(inumber, number)
|
|
if cmplx.IsInf(num) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newStringFormulaArg(cmplx2str(num, value[len(value)-1:]))
|
|
}
|
|
|
|
// IMPRODUCT function calculates the product of two or more complex numbers.
|
|
// The syntax of the function is:
|
|
//
|
|
// IMPRODUCT(number1,[number2],...)
|
|
func (fn *formulaFuncs) IMPRODUCT(argsList *list.List) formulaArg {
|
|
product := complex128(1)
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
if token.Value() == "" {
|
|
continue
|
|
}
|
|
val, err := strconv.ParseComplex(str2cmplx(token.Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
product = product * val
|
|
case ArgNumber:
|
|
product = product * complex(token.Number, 0)
|
|
case ArgMatrix:
|
|
for _, row := range token.Matrix {
|
|
for _, value := range row {
|
|
if value.Value() == "" {
|
|
continue
|
|
}
|
|
val, err := strconv.ParseComplex(str2cmplx(value.Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
product = product * val
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newStringFormulaArg(cmplx2str(product, "i"))
|
|
}
|
|
|
|
// IMREAL function returns the real coefficient of a supplied complex number.
|
|
// The syntax of the function is:
|
|
//
|
|
// IMREAL(inumber)
|
|
func (fn *formulaFuncs) IMREAL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMREAL requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(fmt.Sprint(real(inumber)))
|
|
}
|
|
|
|
// IMSEC function returns the secant of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMSEC(inumber)
|
|
func (fn *formulaFuncs) IMSEC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSEC requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(1/cmplx.Cos(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMSECH function returns the hyperbolic secant of a supplied complex number.
|
|
// The syntax of the function is:
|
|
//
|
|
// IMSECH(inumber)
|
|
func (fn *formulaFuncs) IMSECH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSECH requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(1/cmplx.Cosh(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMSIN function returns the Sine of a supplied complex number. The syntax of
|
|
// the function is:
|
|
//
|
|
// IMSIN(inumber)
|
|
func (fn *formulaFuncs) IMSIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSIN requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Sin(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMSINH function returns the hyperbolic sine of a supplied complex number.
|
|
// The syntax of the function is:
|
|
//
|
|
// IMSINH(inumber)
|
|
func (fn *formulaFuncs) IMSINH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSINH requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Sinh(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMSQRT function returns the square root of a supplied complex number. The
|
|
// syntax of the function is:
|
|
//
|
|
// IMSQRT(inumber)
|
|
func (fn *formulaFuncs) IMSQRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSQRT requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Sqrt(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// IMSUB function calculates the difference between two complex numbers
|
|
// (i.e. subtracts one complex number from another). The syntax of the
|
|
// function is:
|
|
//
|
|
// IMSUB(inumber1,inumber2)
|
|
func (fn *formulaFuncs) IMSUB(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSUB requires 2 arguments")
|
|
}
|
|
i1, err := strconv.ParseComplex(str2cmplx(argsList.Front().Value.(formulaArg).Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
i2, err := strconv.ParseComplex(str2cmplx(argsList.Back().Value.(formulaArg).Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(i1-i2, "i"))
|
|
}
|
|
|
|
// IMSUM function calculates the sum of two or more complex numbers. The
|
|
// syntax of the function is:
|
|
//
|
|
// IMSUM(inumber1,inumber2,...)
|
|
func (fn *formulaFuncs) IMSUM(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMSUM requires at least 1 argument")
|
|
}
|
|
var result complex128
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
num, err := strconv.ParseComplex(str2cmplx(token.Value()), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
result += num
|
|
}
|
|
return newStringFormulaArg(cmplx2str(result, "i"))
|
|
}
|
|
|
|
// IMTAN function returns the tangent of a supplied complex number. The syntax
|
|
// of the function is:
|
|
//
|
|
// IMTAN(inumber)
|
|
func (fn *formulaFuncs) IMTAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IMTAN requires 1 argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).Value()
|
|
inumber, err := strconv.ParseComplex(str2cmplx(value), 128)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorNUM, err.Error())
|
|
}
|
|
return newStringFormulaArg(cmplx2str(cmplx.Tan(inumber), value[len(value)-1:]))
|
|
}
|
|
|
|
// OCT2BIN function converts an Octal (Base 8) number into a Binary (Base 2)
|
|
// number. The syntax of the function is:
|
|
//
|
|
// OCT2BIN(number,[places])
|
|
func (fn *formulaFuncs) OCT2BIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OCT2BIN requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OCT2BIN allows at most 2 arguments")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
decimal, newList := fn.oct2dec(token.Value()), list.New()
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("OCT2BIN", newList)
|
|
}
|
|
|
|
// OCT2DEC function converts an Octal (a base-8 number) into a decimal number.
|
|
// The syntax of the function is:
|
|
//
|
|
// OCT2DEC(number)
|
|
func (fn *formulaFuncs) OCT2DEC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OCT2DEC requires 1 numeric argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
return fn.oct2dec(token.Value())
|
|
}
|
|
|
|
// OCT2HEX function converts an Octal (Base 8) number into a Hexadecimal
|
|
// (Base 16) number. The syntax of the function is:
|
|
//
|
|
// OCT2HEX(number,[places])
|
|
func (fn *formulaFuncs) OCT2HEX(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OCT2HEX requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OCT2HEX allows at most 2 arguments")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
number := token.ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, number.Error)
|
|
}
|
|
decimal, newList := fn.oct2dec(token.Value()), list.New()
|
|
newList.PushBack(decimal)
|
|
if argsList.Len() == 2 {
|
|
newList.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.dec2x("OCT2HEX", newList)
|
|
}
|
|
|
|
// oct2dec is an implementation of the formula function OCT2DEC.
|
|
func (fn *formulaFuncs) oct2dec(number string) formulaArg {
|
|
decimal, length := 0.0, len(number)
|
|
for i := length; i > 0; i-- {
|
|
num, _ := strconv.Atoi(string(number[length-i]))
|
|
if i == 10 && string(number[length-i]) == "7" {
|
|
decimal += math.Pow(-8.0, float64(i-1))
|
|
continue
|
|
}
|
|
decimal += float64(num) * math.Pow(8.0, float64(i-1))
|
|
}
|
|
return newNumberFormulaArg(decimal)
|
|
}
|
|
|
|
// Math and Trigonometric Functions
|
|
|
|
// ABS function returns the absolute value of any supplied number. The syntax
|
|
// of the function is:
|
|
//
|
|
// ABS(number)
|
|
func (fn *formulaFuncs) ABS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ABS requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Abs(arg.Number))
|
|
}
|
|
|
|
// ACOS function calculates the arccosine (i.e. the inverse cosine) of a given
|
|
// number, and returns an angle, in radians, between 0 and π. The syntax of
|
|
// the function is:
|
|
//
|
|
// ACOS(number)
|
|
func (fn *formulaFuncs) ACOS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACOS requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Acos(arg.Number))
|
|
}
|
|
|
|
// ACOSH function calculates the inverse hyperbolic cosine of a supplied number.
|
|
// of the function is:
|
|
//
|
|
// ACOSH(number)
|
|
func (fn *formulaFuncs) ACOSH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACOSH requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Acosh(arg.Number))
|
|
}
|
|
|
|
// ACOT function calculates the arccotangent (i.e. the inverse cotangent) of a
|
|
// given number, and returns an angle, in radians, between 0 and π. The syntax
|
|
// of the function is:
|
|
//
|
|
// ACOT(number)
|
|
func (fn *formulaFuncs) ACOT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACOT requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Pi/2 - math.Atan(arg.Number))
|
|
}
|
|
|
|
// ACOTH function calculates the hyperbolic arccotangent (coth) of a supplied
|
|
// value. The syntax of the function is:
|
|
//
|
|
// ACOTH(number)
|
|
func (fn *formulaFuncs) ACOTH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACOTH requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Atanh(1 / arg.Number))
|
|
}
|
|
|
|
// AGGREGATE function returns the result of a specified operation or function,
|
|
// applied to a list or database of values. The syntax of the function is:
|
|
//
|
|
// AGGREGATE(function_num,options,ref1,[ref2],...)
|
|
func (fn *formulaFuncs) AGGREGATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AGGREGATE requires at least 3 arguments")
|
|
}
|
|
var fnNum, opts formulaArg
|
|
if fnNum = argsList.Front().Value.(formulaArg).ToNumber(); fnNum.Type != ArgNumber {
|
|
return fnNum
|
|
}
|
|
subFn, ok := map[int]func(argsList *list.List) formulaArg{
|
|
1: fn.AVERAGE,
|
|
2: fn.COUNT,
|
|
3: fn.COUNTA,
|
|
4: fn.MAX,
|
|
5: fn.MIN,
|
|
6: fn.PRODUCT,
|
|
7: fn.STDEVdotS,
|
|
8: fn.STDEVdotP,
|
|
9: fn.SUM,
|
|
10: fn.VARdotS,
|
|
11: fn.VARdotP,
|
|
12: fn.MEDIAN,
|
|
13: fn.MODEdotSNGL,
|
|
14: fn.LARGE,
|
|
15: fn.SMALL,
|
|
16: fn.PERCENTILEdotINC,
|
|
17: fn.QUARTILEdotINC,
|
|
18: fn.PERCENTILEdotEXC,
|
|
19: fn.QUARTILEdotEXC,
|
|
}[int(fnNum.Number)]
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AGGREGATE has invalid function_num")
|
|
}
|
|
if opts = argsList.Front().Next().Value.(formulaArg).ToNumber(); opts.Type != ArgNumber {
|
|
return opts
|
|
}
|
|
// TODO: apply option argument values to be ignored during the calculation
|
|
if int(opts.Number) < 0 || int(opts.Number) > 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AGGREGATE has invalid options")
|
|
}
|
|
subArgList := list.New().Init()
|
|
for arg := argsList.Front().Next().Next(); arg != nil; arg = arg.Next() {
|
|
subArgList.PushBack(arg.Value.(formulaArg))
|
|
}
|
|
return subFn(subArgList)
|
|
}
|
|
|
|
// ARABIC function converts a Roman numeral into an Arabic numeral. The syntax
|
|
// of the function is:
|
|
//
|
|
// ARABIC(text)
|
|
func (fn *formulaFuncs) ARABIC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ARABIC requires 1 numeric argument")
|
|
}
|
|
text := argsList.Front().Value.(formulaArg).Value()
|
|
if len(text) > MaxFieldLength {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
text = strings.ToUpper(text)
|
|
number, actualStart, index, isNegative := 0, 0, len(text)-1, false
|
|
startIndex, subtractNumber, currentPartValue, currentCharValue, prevCharValue := 0, 0, 0, 0, -1
|
|
for index >= 0 && text[index] == ' ' {
|
|
index--
|
|
}
|
|
for actualStart <= index && text[actualStart] == ' ' {
|
|
actualStart++
|
|
}
|
|
if actualStart <= index && text[actualStart] == '-' {
|
|
isNegative = true
|
|
actualStart++
|
|
}
|
|
charMap := map[rune]int{'I': 1, 'V': 5, 'X': 10, 'L': 50, 'C': 100, 'D': 500, 'M': 1000}
|
|
for index >= actualStart {
|
|
startIndex = index
|
|
startChar := text[startIndex]
|
|
index--
|
|
for index >= actualStart && (text[index]|' ') == startChar {
|
|
index--
|
|
}
|
|
currentCharValue = charMap[rune(startChar)]
|
|
currentPartValue = (startIndex - index) * currentCharValue
|
|
if currentCharValue >= prevCharValue {
|
|
number += currentPartValue - subtractNumber
|
|
prevCharValue = currentCharValue
|
|
subtractNumber = 0
|
|
continue
|
|
}
|
|
subtractNumber += currentPartValue
|
|
}
|
|
if subtractNumber != 0 {
|
|
number -= subtractNumber
|
|
}
|
|
if isNegative {
|
|
number = -number
|
|
}
|
|
return newNumberFormulaArg(float64(number))
|
|
}
|
|
|
|
// ASIN function calculates the arcsine (i.e. the inverse sine) of a given
|
|
// number, and returns an angle, in radians, between -π/2 and π/2. The syntax
|
|
// of the function is:
|
|
//
|
|
// ASIN(number)
|
|
func (fn *formulaFuncs) ASIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ASIN requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Asin(arg.Number))
|
|
}
|
|
|
|
// ASINH function calculates the inverse hyperbolic sine of a supplied number.
|
|
// The syntax of the function is:
|
|
//
|
|
// ASINH(number)
|
|
func (fn *formulaFuncs) ASINH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ASINH requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Asinh(arg.Number))
|
|
}
|
|
|
|
// ATAN function calculates the arctangent (i.e. the inverse tangent) of a
|
|
// given number, and returns an angle, in radians, between -π/2 and +π/2. The
|
|
// syntax of the function is:
|
|
//
|
|
// ATAN(number)
|
|
func (fn *formulaFuncs) ATAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ATAN requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Atan(arg.Number))
|
|
}
|
|
|
|
// ATANH function calculates the inverse hyperbolic tangent of a supplied
|
|
// number. The syntax of the function is:
|
|
//
|
|
// ATANH(number)
|
|
func (fn *formulaFuncs) ATANH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ATANH requires 1 numeric argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
return newNumberFormulaArg(math.Atanh(arg.Number))
|
|
}
|
|
|
|
// ATAN2 function calculates the arctangent (i.e. the inverse tangent) of a
|
|
// given set of x and y coordinates, and returns an angle, in radians, between
|
|
// -π/2 and +π/2. The syntax of the function is:
|
|
//
|
|
// ATAN2(x_num,y_num)
|
|
func (fn *formulaFuncs) ATAN2(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ATAN2 requires 2 numeric arguments")
|
|
}
|
|
x := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if x.Type == ArgError {
|
|
return x
|
|
}
|
|
y := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if y.Type == ArgError {
|
|
return y
|
|
}
|
|
return newNumberFormulaArg(math.Atan2(x.Number, y.Number))
|
|
}
|
|
|
|
// BASE function converts a number into a supplied base (radix), and returns a
|
|
// text representation of the calculated value. The syntax of the function is:
|
|
//
|
|
// BASE(number,radix,[min_length])
|
|
func (fn *formulaFuncs) BASE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BASE requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BASE allows at most 3 arguments")
|
|
}
|
|
var minLength int
|
|
var err error
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
radix := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if radix.Type == ArgError {
|
|
return radix
|
|
}
|
|
if int(radix.Number) < 2 || int(radix.Number) > 36 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "radix must be an integer >= 2 and <= 36")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
if minLength, err = strconv.Atoi(argsList.Back().Value.(formulaArg).Value()); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
}
|
|
result := strconv.FormatInt(int64(number.Number), int(radix.Number))
|
|
if len(result) < minLength {
|
|
result = strings.Repeat("0", minLength-len(result)) + result
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(result))
|
|
}
|
|
|
|
// CEILING function rounds a supplied number away from zero, to the nearest
|
|
// multiple of a given number. The syntax of the function is:
|
|
//
|
|
// CEILING(number,significance)
|
|
func (fn *formulaFuncs) CEILING(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING allows at most 2 arguments")
|
|
}
|
|
number, significance, res := 0.0, 1.0, 0.0
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
number = n.Number
|
|
if number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
}
|
|
if significance < 0 && number > 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "negative sig to CEILING invalid")
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Ceil(number))
|
|
}
|
|
number, res = math.Modf(number / significance)
|
|
if res > 0 {
|
|
number++
|
|
}
|
|
return newNumberFormulaArg(number * significance)
|
|
}
|
|
|
|
// CEILINGdotMATH function rounds a supplied number up to a supplied multiple
|
|
// of significance. The syntax of the function is:
|
|
//
|
|
// CEILING.MATH(number,[significance],[mode])
|
|
func (fn *formulaFuncs) CEILINGdotMATH(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING.MATH requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING.MATH allows at most 3 arguments")
|
|
}
|
|
number, significance, mode := 0.0, 1.0, 1.0
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
number = n.Number
|
|
if number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Ceil(number))
|
|
}
|
|
if argsList.Len() > 2 {
|
|
m := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if m.Type == ArgError {
|
|
return m
|
|
}
|
|
mode = m.Number
|
|
}
|
|
val, res := math.Modf(number / significance)
|
|
if res != 0 {
|
|
if number > 0 {
|
|
val++
|
|
} else if mode < 0 {
|
|
val--
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val * significance)
|
|
}
|
|
|
|
// CEILINGdotPRECISE function rounds a supplied number up (regardless of the
|
|
// number's sign), to the nearest multiple of a given number. The syntax of
|
|
// the function is:
|
|
//
|
|
// CEILING.PRECISE(number,[significance])
|
|
func (fn *formulaFuncs) CEILINGdotPRECISE(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING.PRECISE requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CEILING.PRECISE allows at most 2 arguments")
|
|
}
|
|
number, significance := 0.0, 1.0
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
number = n.Number
|
|
if number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Ceil(number))
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
significance = math.Abs(significance)
|
|
if significance == 0 {
|
|
return newNumberFormulaArg(significance)
|
|
}
|
|
}
|
|
val, res := math.Modf(number / significance)
|
|
if res != 0 {
|
|
if number > 0 {
|
|
val++
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val * significance)
|
|
}
|
|
|
|
// COMBIN function calculates the number of combinations (in any order) of a
|
|
// given number objects from a set. The syntax of the function is:
|
|
//
|
|
// COMBIN(number,number_chosen)
|
|
func (fn *formulaFuncs) COMBIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMBIN requires 2 argument")
|
|
}
|
|
number, chosen, val := 0.0, 0.0, 1.0
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
number = n.Number
|
|
c := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if c.Type == ArgError {
|
|
return c
|
|
}
|
|
chosen = c.Number
|
|
number, chosen = math.Trunc(number), math.Trunc(chosen)
|
|
if chosen > number {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMBIN requires number >= number_chosen")
|
|
}
|
|
if chosen == number || chosen == 0 {
|
|
return newNumberFormulaArg(1)
|
|
}
|
|
for c := float64(1); c <= chosen; c++ {
|
|
val *= (number + 1 - c) / c
|
|
}
|
|
return newNumberFormulaArg(math.Ceil(val))
|
|
}
|
|
|
|
// COMBINA function calculates the number of combinations, with repetitions,
|
|
// of a given number objects from a set. The syntax of the function is:
|
|
//
|
|
// COMBINA(number,number_chosen)
|
|
func (fn *formulaFuncs) COMBINA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMBINA requires 2 argument")
|
|
}
|
|
var number, chosen float64
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
number = n.Number
|
|
c := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if c.Type == ArgError {
|
|
return c
|
|
}
|
|
chosen = c.Number
|
|
number, chosen = math.Trunc(number), math.Trunc(chosen)
|
|
if number < chosen {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COMBINA requires number > number_chosen")
|
|
}
|
|
if number == 0 {
|
|
return newNumberFormulaArg(number)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(formulaArg{
|
|
String: fmt.Sprintf("%g", number+chosen-1),
|
|
Type: ArgString,
|
|
})
|
|
args.PushBack(formulaArg{
|
|
String: fmt.Sprintf("%g", number-1),
|
|
Type: ArgString,
|
|
})
|
|
return fn.COMBIN(args)
|
|
}
|
|
|
|
// COS function calculates the cosine of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// COS(number)
|
|
func (fn *formulaFuncs) COS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COS requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
return newNumberFormulaArg(math.Cos(val.Number))
|
|
}
|
|
|
|
// COSH function calculates the hyperbolic cosine (cosh) of a supplied number.
|
|
// The syntax of the function is:
|
|
//
|
|
// COSH(number)
|
|
func (fn *formulaFuncs) COSH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COSH requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
return newNumberFormulaArg(math.Cosh(val.Number))
|
|
}
|
|
|
|
// COT function calculates the cotangent of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// COT(number)
|
|
func (fn *formulaFuncs) COT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COT requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(1 / math.Tan(val.Number))
|
|
}
|
|
|
|
// COTH function calculates the hyperbolic cotangent (coth) of a supplied
|
|
// angle. The syntax of the function is:
|
|
//
|
|
// COTH(number)
|
|
func (fn *formulaFuncs) COTH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COTH requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg((math.Exp(val.Number) + math.Exp(-val.Number)) / (math.Exp(val.Number) - math.Exp(-val.Number)))
|
|
}
|
|
|
|
// CSC function calculates the cosecant of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// CSC(number)
|
|
func (fn *formulaFuncs) CSC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CSC requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(1 / math.Sin(val.Number))
|
|
}
|
|
|
|
// CSCH function calculates the hyperbolic cosecant (csch) of a supplied
|
|
// angle. The syntax of the function is:
|
|
//
|
|
// CSCH(number)
|
|
func (fn *formulaFuncs) CSCH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CSCH requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(1 / math.Sinh(val.Number))
|
|
}
|
|
|
|
// DECIMAL function converts a text representation of a number in a specified
|
|
// base, into a decimal value. The syntax of the function is:
|
|
//
|
|
// DECIMAL(text,radix)
|
|
func (fn *formulaFuncs) DECIMAL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DECIMAL requires 2 numeric arguments")
|
|
}
|
|
text := argsList.Front().Value.(formulaArg).Value()
|
|
var err error
|
|
radix := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if radix.Type != ArgNumber {
|
|
return radix
|
|
}
|
|
if len(text) > 2 && (strings.HasPrefix(text, "0x") || strings.HasPrefix(text, "0X")) {
|
|
text = text[2:]
|
|
}
|
|
val, err := strconv.ParseInt(text, int(radix.Number), 64)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
return newNumberFormulaArg(float64(val))
|
|
}
|
|
|
|
// DEGREES function converts radians into degrees. The syntax of the function
|
|
// is:
|
|
//
|
|
// DEGREES(angle)
|
|
func (fn *formulaFuncs) DEGREES(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DEGREES requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(180.0 / math.Pi * val.Number)
|
|
}
|
|
|
|
// EVEN function rounds a supplied number away from zero (i.e. rounds a
|
|
// positive number up and a negative number down), to the next even number.
|
|
// The syntax of the function is:
|
|
//
|
|
// EVEN(number)
|
|
func (fn *formulaFuncs) EVEN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EVEN requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
sign := math.Signbit(number.Number)
|
|
m, frac := math.Modf(number.Number / 2)
|
|
val := m * 2
|
|
if frac != 0 {
|
|
if !sign {
|
|
val += 2
|
|
} else {
|
|
val -= 2
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val)
|
|
}
|
|
|
|
// EXP function calculates the value of the mathematical constant e, raised to
|
|
// the power of a given number. The syntax of the function is:
|
|
//
|
|
// EXP(number)
|
|
func (fn *formulaFuncs) EXP(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EXP requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(fmt.Sprintf("%g", math.Exp(number.Number))))
|
|
}
|
|
|
|
// fact returns the factorial of a supplied number.
|
|
func fact(number float64) float64 {
|
|
val := float64(1)
|
|
for i := float64(2); i <= number; i++ {
|
|
val *= i
|
|
}
|
|
return val
|
|
}
|
|
|
|
// FACT function returns the factorial of a supplied number. The syntax of the
|
|
// function is:
|
|
//
|
|
// FACT(number)
|
|
func (fn *formulaFuncs) FACT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FACT requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(fact(number.Number))
|
|
}
|
|
|
|
// FACTDOUBLE function returns the double factorial of a supplied number. The
|
|
// syntax of the function is:
|
|
//
|
|
// FACTDOUBLE(number)
|
|
func (fn *formulaFuncs) FACTDOUBLE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FACTDOUBLE requires 1 numeric argument")
|
|
}
|
|
val := 1.0
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
for i := math.Trunc(number.Number); i > 1; i -= 2 {
|
|
val *= i
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(fmt.Sprintf("%g", val)))
|
|
}
|
|
|
|
// FLOOR function rounds a supplied number towards zero to the nearest
|
|
// multiple of a specified significance. The syntax of the function is:
|
|
//
|
|
// FLOOR(number,significance)
|
|
func (fn *formulaFuncs) FLOOR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FLOOR requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
significance := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if significance.Type == ArgError {
|
|
return significance
|
|
}
|
|
if significance.Number < 0 && number.Number >= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "invalid arguments to FLOOR")
|
|
}
|
|
val := number.Number
|
|
val, res := math.Modf(val / significance.Number)
|
|
if res != 0 {
|
|
if number.Number < 0 && res < 0 {
|
|
val--
|
|
}
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(fmt.Sprintf("%g", val*significance.Number)))
|
|
}
|
|
|
|
// FLOORdotMATH function rounds a supplied number down to a supplied multiple
|
|
// of significance. The syntax of the function is:
|
|
//
|
|
// FLOOR.MATH(number,[significance],[mode])
|
|
func (fn *formulaFuncs) FLOORdotMATH(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FLOOR.MATH requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FLOOR.MATH allows at most 3 arguments")
|
|
}
|
|
significance, mode := 1.0, 1.0
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Floor(number.Number))
|
|
}
|
|
if argsList.Len() > 2 {
|
|
m := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if m.Type == ArgError {
|
|
return m
|
|
}
|
|
mode = m.Number
|
|
}
|
|
val, res := math.Modf(number.Number / significance)
|
|
if res != 0 && number.Number < 0 && mode > 0 {
|
|
val--
|
|
}
|
|
return newNumberFormulaArg(val * significance)
|
|
}
|
|
|
|
// FLOORdotPRECISE function rounds a supplied number down to a supplied
|
|
// multiple of significance. The syntax of the function is:
|
|
//
|
|
// FLOOR.PRECISE(number,[significance])
|
|
func (fn *formulaFuncs) FLOORdotPRECISE(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FLOOR.PRECISE requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FLOOR.PRECISE allows at most 2 arguments")
|
|
}
|
|
var significance float64
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Floor(number.Number))
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
significance = math.Abs(significance)
|
|
if significance == 0 {
|
|
return newNumberFormulaArg(significance)
|
|
}
|
|
}
|
|
val, res := math.Modf(number.Number / significance)
|
|
if res != 0 {
|
|
if number.Number < 0 {
|
|
val--
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val * significance)
|
|
}
|
|
|
|
// gcd returns the greatest common divisor of two supplied integers.
|
|
func gcd(x, y float64) float64 {
|
|
x, y = math.Trunc(x), math.Trunc(y)
|
|
if x == 0 {
|
|
return y
|
|
}
|
|
if y == 0 {
|
|
return x
|
|
}
|
|
for x != y {
|
|
if x > y {
|
|
x = x - y
|
|
} else {
|
|
y = y - x
|
|
}
|
|
}
|
|
return x
|
|
}
|
|
|
|
// GCD function returns the greatest common divisor of two or more supplied
|
|
// integers. The syntax of the function is:
|
|
//
|
|
// GCD(number1,[number2],...)
|
|
func (fn *formulaFuncs) GCD(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GCD requires at least 1 argument")
|
|
}
|
|
var (
|
|
val float64
|
|
nums []float64
|
|
)
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
num := token.ToNumber()
|
|
if num.Type == ArgError {
|
|
return num
|
|
}
|
|
val = num.Number
|
|
case ArgNumber:
|
|
val = token.Number
|
|
}
|
|
nums = append(nums, val)
|
|
}
|
|
if nums[0] < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GCD only accepts positive arguments")
|
|
}
|
|
if len(nums) == 1 {
|
|
return newNumberFormulaArg(nums[0])
|
|
}
|
|
cd := nums[0]
|
|
for i := 1; i < len(nums); i++ {
|
|
if nums[i] < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GCD only accepts positive arguments")
|
|
}
|
|
cd = gcd(cd, nums[i])
|
|
}
|
|
return newNumberFormulaArg(cd)
|
|
}
|
|
|
|
// INT function truncates a supplied number down to the closest integer. The
|
|
// syntax of the function is:
|
|
//
|
|
// INT(number)
|
|
func (fn *formulaFuncs) INT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "INT requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
val, frac := math.Modf(number.Number)
|
|
if frac < 0 {
|
|
val--
|
|
}
|
|
return newNumberFormulaArg(val)
|
|
}
|
|
|
|
// ISOdotCEILING function rounds a supplied number up (regardless of the
|
|
// number's sign), to the nearest multiple of a supplied significance. The
|
|
// syntax of the function is:
|
|
//
|
|
// ISO.CEILING(number,[significance])
|
|
func (fn *formulaFuncs) ISOdotCEILING(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISO.CEILING requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISO.CEILING allows at most 2 arguments")
|
|
}
|
|
var significance float64
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number < 0 {
|
|
significance = -1
|
|
}
|
|
if argsList.Len() == 1 {
|
|
return newNumberFormulaArg(math.Ceil(number.Number))
|
|
}
|
|
if argsList.Len() > 1 {
|
|
s := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if s.Type == ArgError {
|
|
return s
|
|
}
|
|
significance = s.Number
|
|
significance = math.Abs(significance)
|
|
if significance == 0 {
|
|
return newNumberFormulaArg(significance)
|
|
}
|
|
}
|
|
val, res := math.Modf(number.Number / significance)
|
|
if res != 0 {
|
|
if number.Number > 0 {
|
|
val++
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val * significance)
|
|
}
|
|
|
|
// lcm returns the least common multiple of two supplied integers.
|
|
func lcm(a, b float64) float64 {
|
|
a = math.Trunc(a)
|
|
b = math.Trunc(b)
|
|
if a == 0 && b == 0 {
|
|
return 0
|
|
}
|
|
return a * b / gcd(a, b)
|
|
}
|
|
|
|
// LCM function returns the least common multiple of two or more supplied
|
|
// integers. The syntax of the function is:
|
|
//
|
|
// LCM(number1,[number2],...)
|
|
func (fn *formulaFuncs) LCM(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LCM requires at least 1 argument")
|
|
}
|
|
var (
|
|
val float64
|
|
nums []float64
|
|
err error
|
|
)
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
if token.String == "" {
|
|
continue
|
|
}
|
|
if val, err = strconv.ParseFloat(token.String, 64); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
case ArgNumber:
|
|
val = token.Number
|
|
}
|
|
nums = append(nums, val)
|
|
}
|
|
if nums[0] < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LCM only accepts positive arguments")
|
|
}
|
|
if len(nums) == 1 {
|
|
return newNumberFormulaArg(nums[0])
|
|
}
|
|
cm := nums[0]
|
|
for i := 1; i < len(nums); i++ {
|
|
if nums[i] < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LCM only accepts positive arguments")
|
|
}
|
|
cm = lcm(cm, nums[i])
|
|
}
|
|
return newNumberFormulaArg(cm)
|
|
}
|
|
|
|
// LN function calculates the natural logarithm of a given number. The syntax
|
|
// of the function is:
|
|
//
|
|
// LN(number)
|
|
func (fn *formulaFuncs) LN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LN requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Log(number.Number))
|
|
}
|
|
|
|
// LOG function calculates the logarithm of a given number, to a supplied
|
|
// base. The syntax of the function is:
|
|
//
|
|
// LOG(number,[base])
|
|
func (fn *formulaFuncs) LOG(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOG requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOG allows at most 2 arguments")
|
|
}
|
|
base := 10.0
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if argsList.Len() > 1 {
|
|
b := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if b.Type == ArgError {
|
|
return b
|
|
}
|
|
base = b.Number
|
|
}
|
|
if number.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorDIV)
|
|
}
|
|
if base == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorDIV)
|
|
}
|
|
if base == 1 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(math.Log(number.Number) / math.Log(base))
|
|
}
|
|
|
|
// LOG10 function calculates the base 10 logarithm of a given number. The
|
|
// syntax of the function is:
|
|
//
|
|
// LOG10(number)
|
|
func (fn *formulaFuncs) LOG10(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOG10 requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Log10(number.Number))
|
|
}
|
|
|
|
// minor function implement a minor of a matrix A is the determinant of some
|
|
// smaller square matrix.
|
|
func minor(sqMtx [][]float64, idx int) [][]float64 {
|
|
var ret [][]float64
|
|
for i := range sqMtx {
|
|
if i == 0 {
|
|
continue
|
|
}
|
|
var row []float64
|
|
for j := range sqMtx {
|
|
if j == idx {
|
|
continue
|
|
}
|
|
row = append(row, sqMtx[i][j])
|
|
}
|
|
ret = append(ret, row)
|
|
}
|
|
return ret
|
|
}
|
|
|
|
// det determinant of the 2x2 matrix.
|
|
func det(sqMtx [][]float64) float64 {
|
|
if len(sqMtx) == 2 {
|
|
m00 := sqMtx[0][0]
|
|
m01 := sqMtx[0][1]
|
|
m10 := sqMtx[1][0]
|
|
m11 := sqMtx[1][1]
|
|
return m00*m11 - m10*m01
|
|
}
|
|
var res, sgn float64 = 0, 1
|
|
for j := range sqMtx {
|
|
res += sgn * sqMtx[0][j] * det(minor(sqMtx, j))
|
|
sgn *= -1
|
|
}
|
|
return res
|
|
}
|
|
|
|
// newNumberMatrix converts a formula arguments matrix to a number matrix.
|
|
func newNumberMatrix(arg formulaArg, phalanx bool) (numMtx [][]float64, ele formulaArg) {
|
|
rows := len(arg.Matrix)
|
|
for r, row := range arg.Matrix {
|
|
if phalanx && len(row) != rows {
|
|
ele = newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
return
|
|
}
|
|
numMtx = append(numMtx, make([]float64, len(row)))
|
|
for c, cell := range row {
|
|
if cell.Type != ArgNumber {
|
|
ele = newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
return
|
|
}
|
|
numMtx[r][c] = cell.Number
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// newFormulaArgMatrix converts the number formula arguments matrix to a
|
|
// formula arguments matrix.
|
|
func newFormulaArgMatrix(numMtx [][]float64) (arg [][]formulaArg) {
|
|
for r, row := range numMtx {
|
|
arg = append(arg, make([]formulaArg, len(row)))
|
|
for c, cell := range row {
|
|
arg[r][c] = newNumberFormulaArg(cell)
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// MDETERM calculates the determinant of a square matrix. The
|
|
// syntax of the function is:
|
|
//
|
|
// MDETERM(array)
|
|
func (fn *formulaFuncs) MDETERM(argsList *list.List) (result formulaArg) {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MDETERM requires 1 argument")
|
|
}
|
|
numMtx, errArg := newNumberMatrix(argsList.Front().Value.(formulaArg), true)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
return newNumberFormulaArg(det(numMtx))
|
|
}
|
|
|
|
// cofactorMatrix returns the matrix A of cofactors.
|
|
func cofactorMatrix(i, j int, A [][]float64) float64 {
|
|
N, sign := len(A), -1.0
|
|
if (i+j)%2 == 0 {
|
|
sign = 1
|
|
}
|
|
var B [][]float64
|
|
B = append(B, A...)
|
|
for m := 0; m < N; m++ {
|
|
for n := j + 1; n < N; n++ {
|
|
B[m][n-1] = B[m][n]
|
|
}
|
|
B[m] = B[m][:len(B[m])-1]
|
|
}
|
|
for k := i + 1; k < N; k++ {
|
|
B[k-1] = B[k]
|
|
}
|
|
B = B[:len(B)-1]
|
|
return sign * det(B)
|
|
}
|
|
|
|
// adjugateMatrix returns transpose of the cofactor matrix A with Cramer's
|
|
// rule.
|
|
func adjugateMatrix(A [][]float64) (adjA [][]float64) {
|
|
N := len(A)
|
|
var B [][]float64
|
|
for i := 0; i < N; i++ {
|
|
adjA = append(adjA, make([]float64, N))
|
|
for j := 0; j < N; j++ {
|
|
for m := 0; m < N; m++ {
|
|
for n := 0; n < N; n++ {
|
|
for x := len(B); x <= m; x++ {
|
|
B = append(B, []float64{})
|
|
}
|
|
for k := len(B[m]); k <= n; k++ {
|
|
B[m] = append(B[m], 0)
|
|
}
|
|
B[m][n] = A[m][n]
|
|
}
|
|
}
|
|
adjA[i][j] = cofactorMatrix(j, i, B)
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// MINVERSE function calculates the inverse of a square matrix. The syntax of
|
|
// the function is:
|
|
//
|
|
// MINVERSE(array)
|
|
func (fn *formulaFuncs) MINVERSE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MINVERSE requires 1 argument")
|
|
}
|
|
numMtx, errArg := newNumberMatrix(argsList.Front().Value.(formulaArg), true)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
if detM := det(numMtx); detM != 0 {
|
|
datM, invertM := 1/detM, adjugateMatrix(numMtx)
|
|
for i := 0; i < len(invertM); i++ {
|
|
for j := 0; j < len(invertM[i]); j++ {
|
|
invertM[i][j] *= datM
|
|
}
|
|
}
|
|
return newMatrixFormulaArg(newFormulaArgMatrix(invertM))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
|
|
// MMULT function calculates the matrix product of two arrays
|
|
// (representing matrices). The syntax of the function is:
|
|
//
|
|
// MMULT(array1,array2)
|
|
func (fn *formulaFuncs) MMULT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MMULT requires 2 argument")
|
|
}
|
|
numMtx1, errArg1 := newNumberMatrix(argsList.Front().Value.(formulaArg), false)
|
|
if errArg1.Type == ArgError {
|
|
return errArg1
|
|
}
|
|
numMtx2, errArg2 := newNumberMatrix(argsList.Back().Value.(formulaArg), false)
|
|
if errArg2.Type == ArgError {
|
|
return errArg2
|
|
}
|
|
array2Rows, array2Cols := len(numMtx2), len(numMtx2[0])
|
|
if len(numMtx1[0]) != array2Rows {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
var numMtx [][]float64
|
|
var row1, row []float64
|
|
var sum float64
|
|
for i := 0; i < len(numMtx1); i++ {
|
|
numMtx = append(numMtx, []float64{})
|
|
row = []float64{}
|
|
row1 = numMtx1[i]
|
|
for j := 0; j < array2Cols; j++ {
|
|
sum = 0
|
|
for k := 0; k < array2Rows; k++ {
|
|
sum += row1[k] * numMtx2[k][j]
|
|
}
|
|
for l := len(row); l <= j; l++ {
|
|
row = append(row, 0)
|
|
}
|
|
row[j] = sum
|
|
numMtx[i] = row
|
|
}
|
|
}
|
|
return newMatrixFormulaArg(newFormulaArgMatrix(numMtx))
|
|
}
|
|
|
|
// MOD function returns the remainder of a division between two supplied
|
|
// numbers. The syntax of the function is:
|
|
//
|
|
// MOD(number,divisor)
|
|
func (fn *formulaFuncs) MOD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MOD requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
divisor := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if divisor.Type == ArgError {
|
|
return divisor
|
|
}
|
|
if divisor.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, "MOD divide by zero")
|
|
}
|
|
trunc, rem := math.Modf(number.Number / divisor.Number)
|
|
if rem < 0 {
|
|
trunc--
|
|
}
|
|
return newNumberFormulaArg(number.Number - divisor.Number*trunc)
|
|
}
|
|
|
|
// MROUND function rounds a supplied number up or down to the nearest multiple
|
|
// of a given number. The syntax of the function is:
|
|
//
|
|
// MROUND(number,multiple)
|
|
func (fn *formulaFuncs) MROUND(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MROUND requires 2 numeric arguments")
|
|
}
|
|
n := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if n.Type == ArgError {
|
|
return n
|
|
}
|
|
multiple := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if multiple.Type == ArgError {
|
|
return multiple
|
|
}
|
|
if multiple.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if multiple.Number < 0 && n.Number > 0 ||
|
|
multiple.Number > 0 && n.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
number, res := math.Modf(n.Number / multiple.Number)
|
|
if math.Trunc(res+0.5) > 0 {
|
|
number++
|
|
}
|
|
return newNumberFormulaArg(number * multiple.Number)
|
|
}
|
|
|
|
// MULTINOMIAL function calculates the ratio of the factorial of a sum of
|
|
// supplied values to the product of factorials of those values. The syntax of
|
|
// the function is:
|
|
//
|
|
// MULTINOMIAL(number1,[number2],...)
|
|
func (fn *formulaFuncs) MULTINOMIAL(argsList *list.List) formulaArg {
|
|
val, num, denom := 0.0, 0.0, 1.0
|
|
var err error
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
if token.String == "" {
|
|
continue
|
|
}
|
|
if val, err = strconv.ParseFloat(token.String, 64); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
case ArgNumber:
|
|
val = token.Number
|
|
}
|
|
num += val
|
|
denom *= fact(val)
|
|
}
|
|
return newNumberFormulaArg(fact(num) / denom)
|
|
}
|
|
|
|
// MUNIT function returns the unit matrix for a specified dimension. The
|
|
// syntax of the function is:
|
|
//
|
|
// MUNIT(dimension)
|
|
func (fn *formulaFuncs) MUNIT(argsList *list.List) (result formulaArg) {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MUNIT requires 1 numeric argument")
|
|
}
|
|
dimension := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if dimension.Type == ArgError || dimension.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, dimension.Error)
|
|
}
|
|
matrix := make([][]formulaArg, 0, int(dimension.Number))
|
|
for i := 0; i < int(dimension.Number); i++ {
|
|
row := make([]formulaArg, int(dimension.Number))
|
|
for j := 0; j < int(dimension.Number); j++ {
|
|
if i == j {
|
|
row[j] = newNumberFormulaArg(1.0)
|
|
} else {
|
|
row[j] = newNumberFormulaArg(0.0)
|
|
}
|
|
}
|
|
matrix = append(matrix, row)
|
|
}
|
|
return newMatrixFormulaArg(matrix)
|
|
}
|
|
|
|
// ODD function ounds a supplied number away from zero (i.e. rounds a positive
|
|
// number up and a negative number down), to the next odd number. The syntax
|
|
// of the function is:
|
|
//
|
|
// ODD(number)
|
|
func (fn *formulaFuncs) ODD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ODD requires 1 numeric argument")
|
|
}
|
|
number := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if number.Number == 0 {
|
|
return newNumberFormulaArg(1)
|
|
}
|
|
sign := math.Signbit(number.Number)
|
|
m, frac := math.Modf((number.Number - 1) / 2)
|
|
val := m*2 + 1
|
|
if frac != 0 {
|
|
if !sign {
|
|
val += 2
|
|
} else {
|
|
val -= 2
|
|
}
|
|
}
|
|
return newNumberFormulaArg(val)
|
|
}
|
|
|
|
// PI function returns the value of the mathematical constant π (pi), accurate
|
|
// to 15 digits (14 decimal places). The syntax of the function is:
|
|
//
|
|
// PI()
|
|
func (fn *formulaFuncs) PI(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PI accepts no arguments")
|
|
}
|
|
return newNumberFormulaArg(math.Pi)
|
|
}
|
|
|
|
// POWER function calculates a given number, raised to a supplied power.
|
|
// The syntax of the function is:
|
|
//
|
|
// POWER(number,power)
|
|
func (fn *formulaFuncs) POWER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "POWER requires 2 numeric arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type == ArgError {
|
|
return x
|
|
}
|
|
y := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if y.Type == ArgError {
|
|
return y
|
|
}
|
|
if x.Number == 0 && y.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if x.Number == 0 && y.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(math.Pow(x.Number, y.Number))
|
|
}
|
|
|
|
// PRODUCT function returns the product (multiplication) of a supplied set of
|
|
// numerical values. The syntax of the function is:
|
|
//
|
|
// PRODUCT(number1,[number2],...)
|
|
func (fn *formulaFuncs) PRODUCT(argsList *list.List) formulaArg {
|
|
product := 1.0
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
num := token.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return num
|
|
}
|
|
product = product * num.Number
|
|
case ArgNumber:
|
|
product = product * token.Number
|
|
case ArgMatrix:
|
|
for _, row := range token.Matrix {
|
|
for _, cell := range row {
|
|
if cell.Type == ArgNumber {
|
|
product *= cell.Number
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(product)
|
|
}
|
|
|
|
// QUOTIENT function returns the integer portion of a division between two
|
|
// supplied numbers. The syntax of the function is:
|
|
//
|
|
// QUOTIENT(numerator,denominator)
|
|
func (fn *formulaFuncs) QUOTIENT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "QUOTIENT requires 2 numeric arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type == ArgError {
|
|
return x
|
|
}
|
|
y := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if y.Type == ArgError {
|
|
return y
|
|
}
|
|
if y.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(math.Trunc(x.Number / y.Number))
|
|
}
|
|
|
|
// RADIANS function converts radians into degrees. The syntax of the function is:
|
|
//
|
|
// RADIANS(angle)
|
|
func (fn *formulaFuncs) RADIANS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RADIANS requires 1 numeric argument")
|
|
}
|
|
angle := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if angle.Type == ArgError {
|
|
return angle
|
|
}
|
|
return newNumberFormulaArg(math.Pi / 180.0 * angle.Number)
|
|
}
|
|
|
|
// RAND function generates a random real number between 0 and 1. The syntax of
|
|
// the function is:
|
|
//
|
|
// RAND()
|
|
func (fn *formulaFuncs) RAND(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RAND accepts no arguments")
|
|
}
|
|
return newNumberFormulaArg(rand.New(rand.NewSource(time.Now().UnixNano())).Float64())
|
|
}
|
|
|
|
// RANDBETWEEN function generates a random integer between two supplied
|
|
// integers. The syntax of the function is:
|
|
//
|
|
// RANDBETWEEN(bottom,top)
|
|
func (fn *formulaFuncs) RANDBETWEEN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RANDBETWEEN requires 2 numeric arguments")
|
|
}
|
|
bottom := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if bottom.Type == ArgError {
|
|
return bottom
|
|
}
|
|
top := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if top.Type == ArgError {
|
|
return top
|
|
}
|
|
if top.Number < bottom.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
num := rand.New(rand.NewSource(time.Now().UnixNano())).Int63n(int64(top.Number - bottom.Number + 1))
|
|
return newNumberFormulaArg(float64(num + int64(bottom.Number)))
|
|
}
|
|
|
|
// romanNumerals defined a numeral system that originated in ancient Rome and
|
|
// remained the usual way of writing numbers throughout Europe well into the
|
|
// Late Middle Ages.
|
|
type romanNumerals struct {
|
|
n float64
|
|
s string
|
|
}
|
|
|
|
var romanTable = [][]romanNumerals{
|
|
{
|
|
{1000, "M"},
|
|
{900, "CM"},
|
|
{500, "D"},
|
|
{400, "CD"},
|
|
{100, "C"},
|
|
{90, "XC"},
|
|
{50, "L"},
|
|
{40, "XL"},
|
|
{10, "X"},
|
|
{9, "IX"},
|
|
{5, "V"},
|
|
{4, "IV"},
|
|
{1, "I"},
|
|
},
|
|
{
|
|
{1000, "M"},
|
|
{950, "LM"},
|
|
{900, "CM"},
|
|
{500, "D"},
|
|
{450, "LD"},
|
|
{400, "CD"},
|
|
{100, "C"},
|
|
{95, "VC"},
|
|
{90, "XC"},
|
|
{50, "L"},
|
|
{45, "VL"},
|
|
{40, "XL"},
|
|
{10, "X"},
|
|
{9, "IX"},
|
|
{5, "V"},
|
|
{4, "IV"},
|
|
{1, "I"},
|
|
},
|
|
{
|
|
{1000, "M"},
|
|
{990, "XM"},
|
|
{950, "LM"},
|
|
{900, "CM"},
|
|
{500, "D"},
|
|
{490, "XD"},
|
|
{450, "LD"},
|
|
{400, "CD"},
|
|
{100, "C"},
|
|
{99, "IC"},
|
|
{90, "XC"},
|
|
{50, "L"},
|
|
{45, "VL"},
|
|
{40, "XL"},
|
|
{10, "X"},
|
|
{9, "IX"},
|
|
{5, "V"},
|
|
{4, "IV"},
|
|
{1, "I"},
|
|
},
|
|
{
|
|
{1000, "M"},
|
|
{995, "VM"},
|
|
{990, "XM"},
|
|
{950, "LM"},
|
|
{900, "CM"},
|
|
{500, "D"},
|
|
{495, "VD"},
|
|
{490, "XD"},
|
|
{450, "LD"},
|
|
{400, "CD"},
|
|
{100, "C"},
|
|
{99, "IC"},
|
|
{90, "XC"},
|
|
{50, "L"},
|
|
{45, "VL"},
|
|
{40, "XL"},
|
|
{10, "X"},
|
|
{9, "IX"},
|
|
{5, "V"},
|
|
{4, "IV"},
|
|
{1, "I"},
|
|
},
|
|
{
|
|
{1000, "M"},
|
|
{999, "IM"},
|
|
{995, "VM"},
|
|
{990, "XM"},
|
|
{950, "LM"},
|
|
{900, "CM"},
|
|
{500, "D"},
|
|
{499, "ID"},
|
|
{495, "VD"},
|
|
{490, "XD"},
|
|
{450, "LD"},
|
|
{400, "CD"},
|
|
{100, "C"},
|
|
{99, "IC"},
|
|
{90, "XC"},
|
|
{50, "L"},
|
|
{45, "VL"},
|
|
{40, "XL"},
|
|
{10, "X"},
|
|
{9, "IX"},
|
|
{5, "V"},
|
|
{4, "IV"},
|
|
{1, "I"},
|
|
},
|
|
}
|
|
|
|
// ROMAN function converts an arabic number to Roman. I.e. for a supplied
|
|
// integer, the function returns a text string depicting the roman numeral
|
|
// form of the number. The syntax of the function is:
|
|
//
|
|
// ROMAN(number,[form])
|
|
func (fn *formulaFuncs) ROMAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROMAN requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROMAN allows at most 2 arguments")
|
|
}
|
|
var form int
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if argsList.Len() > 1 {
|
|
f := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if f.Type == ArgError {
|
|
return f
|
|
}
|
|
form = int(f.Number)
|
|
if form < 0 {
|
|
form = 0
|
|
} else if form > 4 {
|
|
form = 4
|
|
}
|
|
}
|
|
decimalTable := romanTable[0]
|
|
switch form {
|
|
case 1:
|
|
decimalTable = romanTable[1]
|
|
case 2:
|
|
decimalTable = romanTable[2]
|
|
case 3:
|
|
decimalTable = romanTable[3]
|
|
case 4:
|
|
decimalTable = romanTable[4]
|
|
}
|
|
val := math.Trunc(number.Number)
|
|
buf := bytes.Buffer{}
|
|
for _, r := range decimalTable {
|
|
for val >= r.n {
|
|
buf.WriteString(r.s)
|
|
val -= r.n
|
|
}
|
|
}
|
|
return newStringFormulaArg(buf.String())
|
|
}
|
|
|
|
type roundMode byte
|
|
|
|
const (
|
|
closest roundMode = iota
|
|
down
|
|
up
|
|
)
|
|
|
|
// round rounds a supplied number up or down.
|
|
func (fn *formulaFuncs) round(number, digits float64, mode roundMode) float64 {
|
|
var significance float64
|
|
if digits > 0 {
|
|
significance = math.Pow(1/10.0, digits)
|
|
} else {
|
|
significance = math.Pow(10.0, -digits)
|
|
}
|
|
val, res := math.Modf(number / significance)
|
|
switch mode {
|
|
case closest:
|
|
const eps = 0.499999999
|
|
if res >= eps {
|
|
val++
|
|
} else if res <= -eps {
|
|
val--
|
|
}
|
|
case down:
|
|
case up:
|
|
if res > 0 {
|
|
val++
|
|
} else if res < 0 {
|
|
val--
|
|
}
|
|
}
|
|
return val * significance
|
|
}
|
|
|
|
// ROUND function rounds a supplied number up or down, to a specified number
|
|
// of decimal places. The syntax of the function is:
|
|
//
|
|
// ROUND(number,num_digits)
|
|
func (fn *formulaFuncs) ROUND(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROUND requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
digits := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if digits.Type == ArgError {
|
|
return digits
|
|
}
|
|
return newNumberFormulaArg(fn.round(number.Number, digits.Number, closest))
|
|
}
|
|
|
|
// ROUNDDOWN function rounds a supplied number down towards zero, to a
|
|
// specified number of decimal places. The syntax of the function is:
|
|
//
|
|
// ROUNDDOWN(number,num_digits)
|
|
func (fn *formulaFuncs) ROUNDDOWN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROUNDDOWN requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
digits := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if digits.Type == ArgError {
|
|
return digits
|
|
}
|
|
return newNumberFormulaArg(fn.round(number.Number, digits.Number, down))
|
|
}
|
|
|
|
// ROUNDUP function rounds a supplied number up, away from zero, to a
|
|
// specified number of decimal places. The syntax of the function is:
|
|
//
|
|
// ROUNDUP(number,num_digits)
|
|
func (fn *formulaFuncs) ROUNDUP(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROUNDUP requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
digits := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if digits.Type == ArgError {
|
|
return digits
|
|
}
|
|
return newNumberFormulaArg(fn.round(number.Number, digits.Number, up))
|
|
}
|
|
|
|
// SEC function calculates the secant of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// SEC(number)
|
|
func (fn *formulaFuncs) SEC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SEC requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Cos(number.Number))
|
|
}
|
|
|
|
// SECH function calculates the hyperbolic secant (sech) of a supplied angle.
|
|
// The syntax of the function is:
|
|
//
|
|
// SECH(number)
|
|
func (fn *formulaFuncs) SECH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SECH requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(1 / math.Cosh(number.Number))
|
|
}
|
|
|
|
// SERIESSUM function returns the sum of a power series. The syntax of the
|
|
// function is:
|
|
//
|
|
// SERIESSUM(x,n,m,coefficients)
|
|
func (fn *formulaFuncs) SERIESSUM(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SERIESSUM requires 4 arguments")
|
|
}
|
|
var x, n, m formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if n = argsList.Front().Next().Value.(formulaArg).ToNumber(); n.Type != ArgNumber {
|
|
return n
|
|
}
|
|
if m = argsList.Front().Next().Next().Value.(formulaArg).ToNumber(); m.Type != ArgNumber {
|
|
return m
|
|
}
|
|
var result, i float64
|
|
for _, coefficient := range argsList.Back().Value.(formulaArg).ToList() {
|
|
if coefficient.Value() == "" {
|
|
continue
|
|
}
|
|
num := coefficient.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return num
|
|
}
|
|
result += num.Number * math.Pow(x.Number, n.Number+(m.Number*i))
|
|
i++
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// SIGN function returns the arithmetic sign (+1, -1 or 0) of a supplied
|
|
// number. I.e. if the number is positive, the Sign function returns +1, if
|
|
// the number is negative, the function returns -1 and if the number is 0
|
|
// (zero), the function returns 0. The syntax of the function is:
|
|
//
|
|
// SIGN(number)
|
|
func (fn *formulaFuncs) SIGN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SIGN requires 1 numeric argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if val.Type == ArgError {
|
|
return val
|
|
}
|
|
if val.Number < 0 {
|
|
return newNumberFormulaArg(-1)
|
|
}
|
|
if val.Number > 0 {
|
|
return newNumberFormulaArg(1)
|
|
}
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
|
|
// SIN function calculates the sine of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// SIN(number)
|
|
func (fn *formulaFuncs) SIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SIN requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Sin(number.Number))
|
|
}
|
|
|
|
// SINH function calculates the hyperbolic sine (sinh) of a supplied number.
|
|
// The syntax of the function is:
|
|
//
|
|
// SINH(number)
|
|
func (fn *formulaFuncs) SINH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SINH requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Sinh(number.Number))
|
|
}
|
|
|
|
// SQRT function calculates the positive square root of a supplied number. The
|
|
// syntax of the function is:
|
|
//
|
|
// SQRT(number)
|
|
func (fn *formulaFuncs) SQRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SQRT requires 1 numeric argument")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if value.Type == ArgError {
|
|
return value
|
|
}
|
|
if value.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(math.Sqrt(value.Number))
|
|
}
|
|
|
|
// SQRTPI function returns the square root of a supplied number multiplied by
|
|
// the mathematical constant, π. The syntax of the function is:
|
|
//
|
|
// SQRTPI(number)
|
|
func (fn *formulaFuncs) SQRTPI(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SQRTPI requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Sqrt(number.Number * math.Pi))
|
|
}
|
|
|
|
// STDEV function calculates the sample standard deviation of a supplied set
|
|
// of values. The syntax of the function is:
|
|
//
|
|
// STDEV(number1,[number2],...)
|
|
func (fn *formulaFuncs) STDEV(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "STDEV requires at least 1 argument")
|
|
}
|
|
return fn.stdev(false, argsList)
|
|
}
|
|
|
|
// STDEVdotS function calculates the sample standard deviation of a supplied
|
|
// set of values. The syntax of the function is:
|
|
//
|
|
// STDEV.S(number1,[number2],...)
|
|
func (fn *formulaFuncs) STDEVdotS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "STDEV.S requires at least 1 argument")
|
|
}
|
|
return fn.stdev(false, argsList)
|
|
}
|
|
|
|
// STDEVA function estimates standard deviation based on a sample. The
|
|
// standard deviation is a measure of how widely values are dispersed from
|
|
// the average value (the mean). The syntax of the function is:
|
|
//
|
|
// STDEVA(number1,[number2],...)
|
|
func (fn *formulaFuncs) STDEVA(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "STDEVA requires at least 1 argument")
|
|
}
|
|
return fn.stdev(true, argsList)
|
|
}
|
|
|
|
// calcStdevPow is part of the implementation stdev.
|
|
func calcStdevPow(result, count float64, n, m formulaArg) (float64, float64) {
|
|
if result == -1 {
|
|
result = math.Pow(n.Number-m.Number, 2)
|
|
} else {
|
|
result += math.Pow(n.Number-m.Number, 2)
|
|
}
|
|
count++
|
|
return result, count
|
|
}
|
|
|
|
// calcStdev is part of the implementation stdev.
|
|
func calcStdev(stdeva bool, result, count float64, mean, token formulaArg) (float64, float64) {
|
|
for _, row := range token.ToList() {
|
|
if row.Type == ArgNumber || row.Type == ArgString {
|
|
if !stdeva && (row.Value() == "TRUE" || row.Value() == "FALSE") {
|
|
continue
|
|
} else if stdeva && (row.Value() == "TRUE" || row.Value() == "FALSE") {
|
|
num := row.ToBool()
|
|
if num.Type == ArgNumber {
|
|
result, count = calcStdevPow(result, count, num, mean)
|
|
continue
|
|
}
|
|
} else {
|
|
num := row.ToNumber()
|
|
if num.Type == ArgNumber {
|
|
result, count = calcStdevPow(result, count, num, mean)
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return result, count
|
|
}
|
|
|
|
// stdev is an implementation of the formula functions STDEV and STDEVA.
|
|
func (fn *formulaFuncs) stdev(stdeva bool, argsList *list.List) formulaArg {
|
|
count, result := -1.0, -1.0
|
|
var mean formulaArg
|
|
if stdeva {
|
|
mean = fn.AVERAGEA(argsList)
|
|
} else {
|
|
mean = fn.AVERAGE(argsList)
|
|
}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString, ArgNumber:
|
|
if !stdeva && (token.Value() == "TRUE" || token.Value() == "FALSE") {
|
|
continue
|
|
} else if stdeva && (token.Value() == "TRUE" || token.Value() == "FALSE") {
|
|
num := token.ToBool()
|
|
if num.Type == ArgNumber {
|
|
result, count = calcStdevPow(result, count, num, mean)
|
|
continue
|
|
}
|
|
} else {
|
|
num := token.ToNumber()
|
|
if num.Type == ArgNumber {
|
|
result, count = calcStdevPow(result, count, num, mean)
|
|
}
|
|
}
|
|
case ArgList, ArgMatrix:
|
|
result, count = calcStdev(stdeva, result, count, mean, token)
|
|
}
|
|
}
|
|
if count > 0 && result >= 0 {
|
|
return newNumberFormulaArg(math.Sqrt(result / count))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
|
|
// POISSONdotDIST function calculates the Poisson Probability Mass Function or
|
|
// the Cumulative Poisson Probability Function for a supplied set of
|
|
// parameters. The syntax of the function is:
|
|
//
|
|
// POISSON.DIST(x,mean,cumulative)
|
|
func (fn *formulaFuncs) POISSONdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "POISSON.DIST requires 3 arguments")
|
|
}
|
|
return fn.POISSON(argsList)
|
|
}
|
|
|
|
// POISSON function calculates the Poisson Probability Mass Function or the
|
|
// Cumulative Poisson Probability Function for a supplied set of parameters.
|
|
// The syntax of the function is:
|
|
//
|
|
// POISSON(x,mean,cumulative)
|
|
func (fn *formulaFuncs) POISSON(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "POISSON requires 3 arguments")
|
|
}
|
|
var x, mean, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if x.Number < 0 || mean.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
summer := 0.0
|
|
floor := math.Floor(x.Number)
|
|
for i := 0; i <= int(floor); i++ {
|
|
summer += math.Pow(mean.Number, float64(i)) / fact(float64(i))
|
|
}
|
|
return newNumberFormulaArg(math.Exp(0-mean.Number) * summer)
|
|
}
|
|
return newNumberFormulaArg(math.Exp(0-mean.Number) * math.Pow(mean.Number, x.Number) / fact(x.Number))
|
|
}
|
|
|
|
// SUBTOTAL function performs a specified calculation (e.g. the sum, product,
|
|
// average, etc.) for a supplied set of values. The syntax of the function is:
|
|
//
|
|
// SUBTOTAL(function_num,ref1,[ref2],...)
|
|
func (fn *formulaFuncs) SUBTOTAL(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUBTOTAL requires at least 2 arguments")
|
|
}
|
|
var fnNum formulaArg
|
|
if fnNum = argsList.Front().Value.(formulaArg).ToNumber(); fnNum.Type != ArgNumber {
|
|
return fnNum
|
|
}
|
|
subFn, ok := map[int]func(argsList *list.List) formulaArg{
|
|
1: fn.AVERAGE, 101: fn.AVERAGE,
|
|
2: fn.COUNT, 102: fn.COUNT,
|
|
3: fn.COUNTA, 103: fn.COUNTA,
|
|
4: fn.MAX, 104: fn.MAX,
|
|
5: fn.MIN, 105: fn.MIN,
|
|
6: fn.PRODUCT, 106: fn.PRODUCT,
|
|
7: fn.STDEV, 107: fn.STDEV,
|
|
8: fn.STDEVP, 108: fn.STDEVP,
|
|
9: fn.SUM, 109: fn.SUM,
|
|
10: fn.VAR, 110: fn.VAR,
|
|
11: fn.VARP, 111: fn.VARP,
|
|
}[int(fnNum.Number)]
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUBTOTAL has invalid function_num")
|
|
}
|
|
subArgList := list.New().Init()
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
subArgList.PushBack(arg.Value.(formulaArg))
|
|
}
|
|
return subFn(subArgList)
|
|
}
|
|
|
|
// SUM function adds together a supplied set of numbers and returns the sum of
|
|
// these values. The syntax of the function is:
|
|
//
|
|
// SUM(number1,[number2],...)
|
|
func (fn *formulaFuncs) SUM(argsList *list.List) formulaArg {
|
|
var sum float64
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgError:
|
|
return token
|
|
case ArgString:
|
|
if num := token.ToNumber(); num.Type == ArgNumber {
|
|
sum += num.Number
|
|
}
|
|
case ArgNumber:
|
|
sum += token.Number
|
|
case ArgMatrix:
|
|
for _, row := range token.Matrix {
|
|
for _, value := range row {
|
|
if num := value.ToNumber(); num.Type == ArgNumber {
|
|
sum += num.Number
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(sum)
|
|
}
|
|
|
|
// SUMIF function finds the values in a supplied array, that satisfy a given
|
|
// criteria, and returns the sum of the corresponding values in a second
|
|
// supplied array. The syntax of the function is:
|
|
//
|
|
// SUMIF(range,criteria,[sum_range])
|
|
func (fn *formulaFuncs) SUMIF(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUMIF requires at least 2 arguments")
|
|
}
|
|
criteria := formulaCriteriaParser(argsList.Front().Next().Value.(formulaArg).String)
|
|
rangeMtx := argsList.Front().Value.(formulaArg).Matrix
|
|
var sumRange [][]formulaArg
|
|
if argsList.Len() == 3 {
|
|
sumRange = argsList.Back().Value.(formulaArg).Matrix
|
|
}
|
|
var sum float64
|
|
var arg formulaArg
|
|
for rowIdx, row := range rangeMtx {
|
|
for colIdx, cell := range row {
|
|
arg = cell
|
|
if arg.Type == ArgEmpty {
|
|
continue
|
|
}
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), criteria); ok {
|
|
if argsList.Len() == 3 {
|
|
if len(sumRange) > rowIdx && len(sumRange[rowIdx]) > colIdx {
|
|
arg = sumRange[rowIdx][colIdx]
|
|
}
|
|
}
|
|
if arg.Type == ArgNumber {
|
|
sum += arg.Number
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(sum)
|
|
}
|
|
|
|
// SUMIFS function finds values in one or more supplied arrays, that satisfy a
|
|
// set of criteria, and returns the sum of the corresponding values in a
|
|
// further supplied array. The syntax of the function is:
|
|
//
|
|
// SUMIFS(sum_range,criteria_range1,criteria1,[criteria_range2,criteria2],...)
|
|
func (fn *formulaFuncs) SUMIFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUMIFS requires at least 3 arguments")
|
|
}
|
|
if argsList.Len()%2 != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var args []formulaArg
|
|
sum, sumRange := 0.0, argsList.Front().Value.(formulaArg).Matrix
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
for _, ref := range formulaIfsMatch(args) {
|
|
if ref.Row >= len(sumRange) || ref.Col >= len(sumRange[ref.Row]) {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if num := sumRange[ref.Row][ref.Col].ToNumber(); num.Type == ArgNumber {
|
|
sum += num.Number
|
|
}
|
|
}
|
|
return newNumberFormulaArg(sum)
|
|
}
|
|
|
|
// sumproduct is an implementation of the formula function SUMPRODUCT.
|
|
func (fn *formulaFuncs) sumproduct(argsList *list.List) formulaArg {
|
|
var (
|
|
argType ArgType
|
|
n int
|
|
res []float64
|
|
sum float64
|
|
)
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
if argType == ArgUnknown {
|
|
argType = token.Type
|
|
}
|
|
if token.Type != argType {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
switch token.Type {
|
|
case ArgString, ArgNumber:
|
|
if num := token.ToNumber(); num.Type == ArgNumber {
|
|
sum = fn.PRODUCT(argsList).Number
|
|
continue
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
case ArgMatrix:
|
|
args := token.ToList()
|
|
if res == nil {
|
|
n = len(args)
|
|
res = make([]float64, n)
|
|
for i := range res {
|
|
res[i] = 1.0
|
|
}
|
|
}
|
|
if len(args) != n {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
for i, value := range args {
|
|
num := value.ToNumber()
|
|
if num.Type != ArgNumber && value.Value() != "" {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
res[i] = res[i] * num.Number
|
|
}
|
|
}
|
|
}
|
|
for _, r := range res {
|
|
sum += r
|
|
}
|
|
return newNumberFormulaArg(sum)
|
|
}
|
|
|
|
// SUMPRODUCT function returns the sum of the products of the corresponding
|
|
// values in a set of supplied arrays. The syntax of the function is:
|
|
//
|
|
// SUMPRODUCT(array1,[array2],[array3],...)
|
|
func (fn *formulaFuncs) SUMPRODUCT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUMPRODUCT requires at least 1 argument")
|
|
}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
if token := arg.Value.(formulaArg); token.Type == ArgError {
|
|
return token
|
|
}
|
|
}
|
|
return fn.sumproduct(argsList)
|
|
}
|
|
|
|
// SUMSQ function returns the sum of squares of a supplied set of values. The
|
|
// syntax of the function is:
|
|
//
|
|
// SUMSQ(number1,[number2],...)
|
|
func (fn *formulaFuncs) SUMSQ(argsList *list.List) formulaArg {
|
|
var val, sq float64
|
|
var err error
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
if token.String == "" {
|
|
continue
|
|
}
|
|
if val, err = strconv.ParseFloat(token.String, 64); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
sq += val * val
|
|
case ArgNumber:
|
|
sq += token.Number * token.Number
|
|
case ArgMatrix:
|
|
for _, row := range token.Matrix {
|
|
for _, value := range row {
|
|
if value.Value() == "" {
|
|
continue
|
|
}
|
|
if val, err = strconv.ParseFloat(value.Value(), 64); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
sq += val * val
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(sq)
|
|
}
|
|
|
|
// sumx is an implementation of the formula functions SUMX2MY2, SUMX2PY2 and
|
|
// SUMXMY2.
|
|
func (fn *formulaFuncs) sumx(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 arguments", name))
|
|
}
|
|
array1 := argsList.Front().Value.(formulaArg)
|
|
array2 := argsList.Back().Value.(formulaArg)
|
|
left, right := array1.ToList(), array2.ToList()
|
|
n := len(left)
|
|
if n != len(right) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
result := 0.0
|
|
for i := 0; i < n; i++ {
|
|
if lhs, rhs := left[i].ToNumber(), right[i].ToNumber(); lhs.Number != 0 && rhs.Number != 0 {
|
|
switch name {
|
|
case "SUMX2MY2":
|
|
result += lhs.Number*lhs.Number - rhs.Number*rhs.Number
|
|
case "SUMX2PY2":
|
|
result += lhs.Number*lhs.Number + rhs.Number*rhs.Number
|
|
default:
|
|
result += (lhs.Number - rhs.Number) * (lhs.Number - rhs.Number)
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// SUMX2MY2 function returns the sum of the differences of squares of two
|
|
// supplied sets of values. The syntax of the function is:
|
|
//
|
|
// SUMX2MY2(array_x,array_y)
|
|
func (fn *formulaFuncs) SUMX2MY2(argsList *list.List) formulaArg {
|
|
return fn.sumx("SUMX2MY2", argsList)
|
|
}
|
|
|
|
// SUMX2PY2 function returns the sum of the sum of squares of two supplied sets
|
|
// of values. The syntax of the function is:
|
|
//
|
|
// SUMX2PY2(array_x,array_y)
|
|
func (fn *formulaFuncs) SUMX2PY2(argsList *list.List) formulaArg {
|
|
return fn.sumx("SUMX2PY2", argsList)
|
|
}
|
|
|
|
// SUMXMY2 function returns the sum of the squares of differences between
|
|
// corresponding values in two supplied arrays. The syntax of the function
|
|
// is:
|
|
//
|
|
// SUMXMY2(array_x,array_y)
|
|
func (fn *formulaFuncs) SUMXMY2(argsList *list.List) formulaArg {
|
|
return fn.sumx("SUMXMY2", argsList)
|
|
}
|
|
|
|
// TAN function calculates the tangent of a given angle. The syntax of the
|
|
// function is:
|
|
//
|
|
// TAN(number)
|
|
func (fn *formulaFuncs) TAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TAN requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Tan(number.Number))
|
|
}
|
|
|
|
// TANH function calculates the hyperbolic tangent (tanh) of a supplied
|
|
// number. The syntax of the function is:
|
|
//
|
|
// TANH(number)
|
|
func (fn *formulaFuncs) TANH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TANH requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
return newNumberFormulaArg(math.Tanh(number.Number))
|
|
}
|
|
|
|
// TRUNC function truncates a supplied number to a specified number of decimal
|
|
// places. The syntax of the function is:
|
|
//
|
|
// TRUNC(number,[number_digits])
|
|
func (fn *formulaFuncs) TRUNC(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TRUNC requires at least 1 argument")
|
|
}
|
|
var digits, adjust, rtrim float64
|
|
var err error
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type == ArgError {
|
|
return number
|
|
}
|
|
if argsList.Len() > 1 {
|
|
d := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if d.Type == ArgError {
|
|
return d
|
|
}
|
|
digits = d.Number
|
|
digits = math.Floor(digits)
|
|
}
|
|
adjust = math.Pow(10, digits)
|
|
x := int((math.Abs(number.Number) - math.Abs(float64(int(number.Number)))) * adjust)
|
|
if x != 0 {
|
|
if rtrim, err = strconv.ParseFloat(strings.TrimRight(strconv.Itoa(x), "0"), 64); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
}
|
|
if (digits > 0) && (rtrim < adjust/10) {
|
|
return newNumberFormulaArg(number.Number)
|
|
}
|
|
return newNumberFormulaArg(float64(int(number.Number*adjust)) / adjust)
|
|
}
|
|
|
|
// Statistical Functions
|
|
|
|
// AVEDEV function calculates the average deviation of a supplied set of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// AVEDEV(number1,[number2],...)
|
|
func (fn *formulaFuncs) AVEDEV(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AVEDEV requires at least 1 argument")
|
|
}
|
|
average := fn.AVERAGE(argsList)
|
|
if average.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
result, count := 0.0, 0.0
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
num := arg.Value.(formulaArg).ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
result += math.Abs(num.Number - average.Number)
|
|
count++
|
|
}
|
|
return newNumberFormulaArg(result / count)
|
|
}
|
|
|
|
// AVERAGE function returns the arithmetic mean of a list of supplied numbers.
|
|
// The syntax of the function is:
|
|
//
|
|
// AVERAGE(number1,[number2],...)
|
|
func (fn *formulaFuncs) AVERAGE(argsList *list.List) formulaArg {
|
|
var args []formulaArg
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
count, sum := fn.countSum(false, args)
|
|
if count == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(sum / count)
|
|
}
|
|
|
|
// AVERAGEA function returns the arithmetic mean of a list of supplied numbers
|
|
// with text cell and zero values. The syntax of the function is:
|
|
//
|
|
// AVERAGEA(number1,[number2],...)
|
|
func (fn *formulaFuncs) AVERAGEA(argsList *list.List) formulaArg {
|
|
var args []formulaArg
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
count, sum := fn.countSum(true, args)
|
|
if count == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(sum / count)
|
|
}
|
|
|
|
// AVERAGEIF function finds the values in a supplied array that satisfy a
|
|
// specified criteria, and returns the average (i.e. the statistical mean) of
|
|
// the corresponding values in a second supplied array. The syntax of the
|
|
// function is:
|
|
//
|
|
// AVERAGEIF(range,criteria,[average_range])
|
|
func (fn *formulaFuncs) AVERAGEIF(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AVERAGEIF requires at least 2 arguments")
|
|
}
|
|
var (
|
|
criteria = formulaCriteriaParser(argsList.Front().Next().Value.(formulaArg).Value())
|
|
rangeMtx = argsList.Front().Value.(formulaArg).Matrix
|
|
cellRange [][]formulaArg
|
|
args []formulaArg
|
|
val float64
|
|
err error
|
|
ok bool
|
|
)
|
|
if argsList.Len() == 3 {
|
|
cellRange = argsList.Back().Value.(formulaArg).Matrix
|
|
}
|
|
for rowIdx, row := range rangeMtx {
|
|
for colIdx, col := range row {
|
|
fromVal := col.Value()
|
|
if col.Value() == "" {
|
|
continue
|
|
}
|
|
ok, _ = formulaCriteriaEval(fromVal, criteria)
|
|
if ok {
|
|
if argsList.Len() == 3 {
|
|
if len(cellRange) > rowIdx && len(cellRange[rowIdx]) > colIdx {
|
|
fromVal = cellRange[rowIdx][colIdx].Value()
|
|
}
|
|
}
|
|
if val, err = strconv.ParseFloat(fromVal, 64); err != nil {
|
|
continue
|
|
}
|
|
args = append(args, newNumberFormulaArg(val))
|
|
}
|
|
}
|
|
}
|
|
count, sum := fn.countSum(false, args)
|
|
if count == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(sum / count)
|
|
}
|
|
|
|
// AVERAGEIFS function finds entries in one or more arrays, that satisfy a set
|
|
// of supplied criteria, and returns the average (i.e. the statistical mean)
|
|
// of the corresponding values in a further supplied array. The syntax of the
|
|
// function is:
|
|
//
|
|
// AVERAGEIFS(average_range,criteria_range1,criteria1,[criteria_range2,criteria2],...)
|
|
func (fn *formulaFuncs) AVERAGEIFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AVERAGEIFS requires at least 3 arguments")
|
|
}
|
|
if argsList.Len()%2 != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var args []formulaArg
|
|
sum, sumRange := 0.0, argsList.Front().Value.(formulaArg).Matrix
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
count := 0.0
|
|
for _, ref := range formulaIfsMatch(args) {
|
|
if num := sumRange[ref.Row][ref.Col].ToNumber(); num.Type == ArgNumber {
|
|
sum += num.Number
|
|
count++
|
|
}
|
|
}
|
|
if count == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, "AVERAGEIF divide by zero")
|
|
}
|
|
return newNumberFormulaArg(sum / count)
|
|
}
|
|
|
|
// getBetaHelperContFrac continued fractions for the beta function.
|
|
func getBetaHelperContFrac(fX, fA, fB float64) float64 {
|
|
var a1, b1, a2, b2, fnorm, cfnew, cf, rm float64
|
|
a1, b1, b2 = 1, 1, 1-(fA+fB)/(fA+1)*fX
|
|
if b2 == 0 {
|
|
a2, fnorm, cf = 0, 1, 1
|
|
} else {
|
|
a2, fnorm = 1, 1/b2
|
|
cf = a2 * fnorm
|
|
}
|
|
cfnew, rm = 1, 1
|
|
fMaxIter, fMachEps := 50000.0, 2.22045e-016
|
|
bfinished := false
|
|
for rm < fMaxIter && !bfinished {
|
|
apl2m := fA + 2*rm
|
|
d2m := rm * (fB - rm) * fX / ((apl2m - 1) * apl2m)
|
|
d2m1 := -(fA + rm) * (fA + fB + rm) * fX / (apl2m * (apl2m + 1))
|
|
a1 = (a2 + d2m*a1) * fnorm
|
|
b1 = (b2 + d2m*b1) * fnorm
|
|
a2 = a1 + d2m1*a2*fnorm
|
|
b2 = b1 + d2m1*b2*fnorm
|
|
if b2 != 0 {
|
|
fnorm = 1 / b2
|
|
cfnew = a2 * fnorm
|
|
bfinished = math.Abs(cf-cfnew) < math.Abs(cf)*fMachEps
|
|
}
|
|
cf = cfnew
|
|
rm++
|
|
}
|
|
return cf
|
|
}
|
|
|
|
// getLanczosSum uses a variant of the Lanczos sum with a rational function.
|
|
func getLanczosSum(fZ float64) float64 {
|
|
num := []float64{
|
|
23531376880.41075968857200767445163675473,
|
|
42919803642.64909876895789904700198885093,
|
|
35711959237.35566804944018545154716670596,
|
|
17921034426.03720969991975575445893111267,
|
|
6039542586.35202800506429164430729792107,
|
|
1439720407.311721673663223072794912393972,
|
|
248874557.8620541565114603864132294232163,
|
|
31426415.58540019438061423162831820536287,
|
|
2876370.628935372441225409051620849613599,
|
|
186056.2653952234950402949897160456992822,
|
|
8071.672002365816210638002902272250613822,
|
|
210.8242777515793458725097339207133627117,
|
|
2.506628274631000270164908177133837338626,
|
|
}
|
|
denom := []float64{
|
|
0,
|
|
39916800,
|
|
120543840,
|
|
150917976,
|
|
105258076,
|
|
45995730,
|
|
13339535,
|
|
2637558,
|
|
357423,
|
|
32670,
|
|
1925,
|
|
66,
|
|
1,
|
|
}
|
|
var sumNum, sumDenom, zInv float64
|
|
if fZ <= 1 {
|
|
sumNum = num[12]
|
|
sumDenom = denom[12]
|
|
for i := 11; i >= 0; i-- {
|
|
sumNum *= fZ
|
|
sumNum += num[i]
|
|
sumDenom *= fZ
|
|
sumDenom += denom[i]
|
|
}
|
|
} else {
|
|
zInv = 1 / fZ
|
|
sumNum = num[0]
|
|
sumDenom = denom[0]
|
|
for i := 1; i <= 12; i++ {
|
|
sumNum *= zInv
|
|
sumNum += num[i]
|
|
sumDenom *= zInv
|
|
sumDenom += denom[i]
|
|
}
|
|
}
|
|
return sumNum / sumDenom
|
|
}
|
|
|
|
// getBeta return beta distribution.
|
|
func getBeta(fAlpha, fBeta float64) float64 {
|
|
var fA, fB float64
|
|
if fAlpha > fBeta {
|
|
fA = fAlpha
|
|
fB = fBeta
|
|
} else {
|
|
fA = fBeta
|
|
fB = fAlpha
|
|
}
|
|
const maxGammaArgument = 171.624376956302
|
|
if fA+fB < maxGammaArgument {
|
|
return math.Gamma(fA) / math.Gamma(fA+fB) * math.Gamma(fB)
|
|
}
|
|
fg := 6.024680040776729583740234375
|
|
fgm := fg - 0.5
|
|
fLanczos := getLanczosSum(fA)
|
|
fLanczos /= getLanczosSum(fA + fB)
|
|
fLanczos *= getLanczosSum(fB)
|
|
fABgm := fA + fB + fgm
|
|
fLanczos *= math.Sqrt((fABgm / (fA + fgm)) / (fB + fgm))
|
|
fTempA := fB / (fA + fgm)
|
|
fTempB := fA / (fB + fgm)
|
|
fResult := math.Exp(-fA*math.Log1p(fTempA) - fB*math.Log1p(fTempB) - fgm)
|
|
fResult *= fLanczos
|
|
return fResult
|
|
}
|
|
|
|
// getBetaDistPDF is an implementation for the Beta probability density
|
|
// function.
|
|
func getBetaDistPDF(fX, fA, fB float64) float64 {
|
|
if fX <= 0 || fX >= 1 {
|
|
return 0
|
|
}
|
|
fLogDblMax, fLogDblMin := math.Log(1.79769e+308), math.Log(2.22507e-308)
|
|
fLogY := math.Log(0.5 - fX + 0.5)
|
|
if fX < 0.1 {
|
|
fLogY = math.Log1p(-fX)
|
|
}
|
|
fLogX := math.Log(fX)
|
|
fAm1LogX := (fA - 1) * fLogX
|
|
fBm1LogY := (fB - 1) * fLogY
|
|
fLogBeta := getLogBeta(fA, fB)
|
|
if fAm1LogX < fLogDblMax && fAm1LogX > fLogDblMin && fBm1LogY < fLogDblMax &&
|
|
fBm1LogY > fLogDblMin && fLogBeta < fLogDblMax && fLogBeta > fLogDblMin &&
|
|
fAm1LogX+fBm1LogY < fLogDblMax && fAm1LogX+fBm1LogY > fLogDblMin {
|
|
return math.Pow(fX, fA-1) * math.Pow(0.5-fX+0.5, fB-1) / getBeta(fA, fB)
|
|
}
|
|
return math.Exp(fAm1LogX + fBm1LogY - fLogBeta)
|
|
}
|
|
|
|
// getLogBeta return beta with logarithm.
|
|
func getLogBeta(fAlpha, fBeta float64) float64 {
|
|
var fA, fB float64
|
|
if fAlpha > fBeta {
|
|
fA, fB = fAlpha, fBeta
|
|
} else {
|
|
fA, fB = fBeta, fAlpha
|
|
}
|
|
fg := 6.024680040776729583740234375
|
|
fgm := fg - 0.5
|
|
fLanczos := getLanczosSum(fA)
|
|
fLanczos /= getLanczosSum(fA + fB)
|
|
fLanczos *= getLanczosSum(fB)
|
|
fLogLanczos := math.Log(fLanczos)
|
|
fABgm := fA + fB + fgm
|
|
fLogLanczos += 0.5 * (math.Log(fABgm) - math.Log(fA+fgm) - math.Log(fB+fgm))
|
|
fTempA := fB / (fA + fgm)
|
|
fTempB := fA / (fB + fgm)
|
|
fResult := -fA*math.Log1p(fTempA) - fB*math.Log1p(fTempB) - fgm
|
|
fResult += fLogLanczos
|
|
return fResult
|
|
}
|
|
|
|
// getBetaDist is an implementation for the beta distribution function.
|
|
func getBetaDist(fXin, fAlpha, fBeta float64) float64 {
|
|
if fXin <= 0 {
|
|
return 0
|
|
}
|
|
if fXin >= 1 {
|
|
return 1
|
|
}
|
|
if fBeta == 1 {
|
|
return math.Pow(fXin, fAlpha)
|
|
}
|
|
if fAlpha == 1 {
|
|
return -math.Expm1(fBeta * math.Log1p(-fXin))
|
|
}
|
|
var fResult float64
|
|
fY, flnY := (0.5-fXin)+0.5, math.Log1p(-fXin)
|
|
fX, flnX := fXin, math.Log(fXin)
|
|
fA, fB := fAlpha, fBeta
|
|
bReflect := fXin > fAlpha/(fAlpha+fBeta)
|
|
if bReflect {
|
|
fA = fBeta
|
|
fB = fAlpha
|
|
fX = fY
|
|
fY = fXin
|
|
flnX = flnY
|
|
flnY = math.Log(fXin)
|
|
}
|
|
fResult = getBetaHelperContFrac(fX, fA, fB) / fA
|
|
fP, fQ := fA/(fA+fB), fB/(fA+fB)
|
|
var fTemp float64
|
|
if fA > 1 && fB > 1 && fP < 0.97 && fQ < 0.97 {
|
|
fTemp = getBetaDistPDF(fX, fA, fB) * fX * fY
|
|
} else {
|
|
fTemp = math.Exp(fA*flnX + fB*flnY - getLogBeta(fA, fB))
|
|
}
|
|
fResult *= fTemp
|
|
if bReflect {
|
|
fResult = 0.5 - fResult + 0.5
|
|
}
|
|
return fResult
|
|
}
|
|
|
|
// prepareBETAdotDISTArgs checking and prepare arguments for the formula
|
|
// function BETA.DIST.
|
|
func (fn *formulaFuncs) prepareBETAdotDISTArgs(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BETA.DIST requires at least 4 arguments")
|
|
}
|
|
if argsList.Len() > 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BETA.DIST requires at most 6 arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
alpha := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
beta := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if beta.Type != ArgNumber {
|
|
return beta
|
|
}
|
|
if alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
cumulative := argsList.Front().Next().Next().Next().Value.(formulaArg).ToBool()
|
|
if cumulative.Type != ArgNumber {
|
|
return cumulative
|
|
}
|
|
a, b := newNumberFormulaArg(0), newNumberFormulaArg(1)
|
|
if argsList.Len() > 4 {
|
|
if a = argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); a.Type != ArgNumber {
|
|
return a
|
|
}
|
|
}
|
|
if argsList.Len() == 6 {
|
|
if b = argsList.Back().Value.(formulaArg).ToNumber(); b.Type != ArgNumber {
|
|
return b
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{x, alpha, beta, cumulative, a, b})
|
|
}
|
|
|
|
// BETAdotDIST function calculates the cumulative beta distribution function
|
|
// or the probability density function of the Beta distribution, for a
|
|
// supplied set of parameters. The syntax of the function is:
|
|
//
|
|
// BETA.DIST(x,alpha,beta,cumulative,[A],[B])
|
|
func (fn *formulaFuncs) BETAdotDIST(argsList *list.List) formulaArg {
|
|
args := fn.prepareBETAdotDISTArgs(argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
x, alpha, beta, cumulative, a, b := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5]
|
|
if x.Number < a.Number || x.Number > b.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if a.Number == b.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
scale := b.Number - a.Number
|
|
x.Number = (x.Number - a.Number) / scale
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(getBetaDist(x.Number, alpha.Number, beta.Number))
|
|
}
|
|
return newNumberFormulaArg(getBetaDistPDF(x.Number, alpha.Number, beta.Number) / scale)
|
|
}
|
|
|
|
// BETADIST function calculates the cumulative beta probability density
|
|
// function for a supplied set of parameters. The syntax of the function is:
|
|
//
|
|
// BETADIST(x,alpha,beta,[A],[B])
|
|
func (fn *formulaFuncs) BETADIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BETADIST requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BETADIST requires at most 5 arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
alpha := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
beta := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if beta.Type != ArgNumber {
|
|
return beta
|
|
}
|
|
if alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
a, b := newNumberFormulaArg(0), newNumberFormulaArg(1)
|
|
if argsList.Len() > 3 {
|
|
if a = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); a.Type != ArgNumber {
|
|
return a
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if b = argsList.Back().Value.(formulaArg).ToNumber(); b.Type != ArgNumber {
|
|
return b
|
|
}
|
|
}
|
|
if x.Number < a.Number || x.Number > b.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if a.Number == b.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(getBetaDist((x.Number-a.Number)/(b.Number-a.Number), alpha.Number, beta.Number))
|
|
}
|
|
|
|
// d1mach returns double precision real machine constants.
|
|
func d1mach(i int) float64 {
|
|
arr := []float64{
|
|
2.2250738585072014e-308,
|
|
1.7976931348623158e+308,
|
|
1.1102230246251565e-16,
|
|
2.2204460492503131e-16,
|
|
0.301029995663981195,
|
|
}
|
|
if i > len(arr) {
|
|
return 0
|
|
}
|
|
return arr[i-1]
|
|
}
|
|
|
|
// chebyshevInit determines the number of terms for the double precision
|
|
// orthogonal series "dos" needed to insure the error is no larger
|
|
// than "eta". Ordinarily eta will be chosen to be one-tenth machine
|
|
// precision.
|
|
func chebyshevInit(nos int, eta float64, dos []float64) int {
|
|
i, e := 0, 0.0
|
|
if nos < 1 {
|
|
return 0
|
|
}
|
|
for ii := 1; ii <= nos; ii++ {
|
|
i = nos - ii
|
|
e += math.Abs(dos[i])
|
|
if e > eta {
|
|
return i
|
|
}
|
|
}
|
|
return i
|
|
}
|
|
|
|
// chebyshevEval evaluates the n-term Chebyshev series "a" at "x".
|
|
func chebyshevEval(n int, x float64, a []float64) float64 {
|
|
if n < 1 || n > 1000 || x < -1.1 || x > 1.1 {
|
|
return math.NaN()
|
|
}
|
|
twox, b0, b1, b2 := x*2, 0.0, 0.0, 0.0
|
|
for i := 1; i <= n; i++ {
|
|
b2 = b1
|
|
b1 = b0
|
|
b0 = twox*b1 - b2 + a[n-i]
|
|
}
|
|
return (b0 - b2) * 0.5
|
|
}
|
|
|
|
// lgammacor is an implementation for the log(gamma) correction.
|
|
func lgammacor(x float64) float64 {
|
|
algmcs := []float64{
|
|
0.1666389480451863247205729650822, -0.1384948176067563840732986059135e-4,
|
|
0.9810825646924729426157171547487e-8, -0.1809129475572494194263306266719e-10,
|
|
0.6221098041892605227126015543416e-13, -0.3399615005417721944303330599666e-15,
|
|
0.2683181998482698748957538846666e-17, -0.2868042435334643284144622399999e-19,
|
|
0.3962837061046434803679306666666e-21, -0.6831888753985766870111999999999e-23,
|
|
0.1429227355942498147573333333333e-24, -0.3547598158101070547199999999999e-26,
|
|
0.1025680058010470912000000000000e-27, -0.3401102254316748799999999999999e-29,
|
|
0.1276642195630062933333333333333e-30,
|
|
}
|
|
nalgm := chebyshevInit(15, d1mach(3), algmcs)
|
|
xbig := 1.0 / math.Sqrt(d1mach(3))
|
|
xmax := math.Exp(math.Min(math.Log(d1mach(2)/12.0), -math.Log(12.0*d1mach(1))))
|
|
if x < 10.0 {
|
|
return math.NaN()
|
|
} else if x >= xmax {
|
|
return 4.930380657631324e-32
|
|
} else if x < xbig {
|
|
tmp := 10.0 / x
|
|
return chebyshevEval(nalgm, tmp*tmp*2.0-1.0, algmcs) / x
|
|
}
|
|
return 1.0 / (x * 12.0)
|
|
}
|
|
|
|
// logrelerr compute the relative error logarithm.
|
|
func logrelerr(x float64) float64 {
|
|
alnrcs := []float64{
|
|
0.10378693562743769800686267719098e+1, -0.13364301504908918098766041553133,
|
|
0.19408249135520563357926199374750e-1, -0.30107551127535777690376537776592e-2,
|
|
0.48694614797154850090456366509137e-3, -0.81054881893175356066809943008622e-4,
|
|
0.13778847799559524782938251496059e-4, -0.23802210894358970251369992914935e-5,
|
|
0.41640416213865183476391859901989e-6, -0.73595828378075994984266837031998e-7,
|
|
0.13117611876241674949152294345011e-7, -0.23546709317742425136696092330175e-8,
|
|
0.42522773276034997775638052962567e-9, -0.77190894134840796826108107493300e-10,
|
|
0.14075746481359069909215356472191e-10, -0.25769072058024680627537078627584e-11,
|
|
0.47342406666294421849154395005938e-12, -0.87249012674742641745301263292675e-13,
|
|
0.16124614902740551465739833119115e-13, -0.29875652015665773006710792416815e-14,
|
|
0.55480701209082887983041321697279e-15, -0.10324619158271569595141333961932e-15,
|
|
0.19250239203049851177878503244868e-16, -0.35955073465265150011189707844266e-17,
|
|
0.67264542537876857892194574226773e-18, -0.12602624168735219252082425637546e-18,
|
|
0.23644884408606210044916158955519e-19, -0.44419377050807936898878389179733e-20,
|
|
0.83546594464034259016241293994666e-21, -0.15731559416479562574899253521066e-21,
|
|
0.29653128740247422686154369706666e-22, -0.55949583481815947292156013226666e-23,
|
|
0.10566354268835681048187284138666e-23, -0.19972483680670204548314999466666e-24,
|
|
0.37782977818839361421049855999999e-25, -0.71531586889081740345038165333333e-26,
|
|
0.13552488463674213646502024533333e-26, -0.25694673048487567430079829333333e-27,
|
|
0.48747756066216949076459519999999e-28, -0.92542112530849715321132373333333e-29,
|
|
0.17578597841760239233269760000000e-29, -0.33410026677731010351377066666666e-30,
|
|
0.63533936180236187354180266666666e-31,
|
|
}
|
|
nlnrel := chebyshevInit(43, 0.1*d1mach(3), alnrcs)
|
|
if x <= -1 {
|
|
return math.NaN()
|
|
}
|
|
if math.Abs(x) <= 0.375 {
|
|
return x * (1.0 - x*chebyshevEval(nlnrel, x/0.375, alnrcs))
|
|
}
|
|
return math.Log(x + 1.0)
|
|
}
|
|
|
|
// logBeta is an implementation for the log of the beta distribution
|
|
// function.
|
|
func logBeta(a, b float64) float64 {
|
|
corr, p, q := 0.0, a, a
|
|
if b < p {
|
|
p = b
|
|
}
|
|
if b > q {
|
|
q = b
|
|
}
|
|
if p < 0 {
|
|
return math.NaN()
|
|
}
|
|
if p == 0 {
|
|
return math.MaxFloat64
|
|
}
|
|
if p >= 10.0 {
|
|
corr = lgammacor(p) + lgammacor(q) - lgammacor(p+q)
|
|
f1 := q * logrelerr(-p/(p+q))
|
|
return math.Log(q)*-0.5 + 0.918938533204672741780329736406 + corr + (p-0.5)*math.Log(p/(p+q)) + math.Nextafter(f1, f1)
|
|
}
|
|
if q >= 10 {
|
|
corr = lgammacor(q) - lgammacor(p+q)
|
|
val, _ := math.Lgamma(p)
|
|
return val + corr + p - p*math.Log(p+q) + (q-0.5)*logrelerr(-p/(p+q))
|
|
}
|
|
return math.Log(math.Gamma(p) * (math.Gamma(q) / math.Gamma(p+q)))
|
|
}
|
|
|
|
// pbetaRaw is a part of pbeta for the beta distribution.
|
|
func pbetaRaw(alnsml, ans, eps, p, pin, q, sml, x, y float64) float64 {
|
|
if q > 1.0 {
|
|
xb := p*math.Log(y) + q*math.Log(1.0-y) - logBeta(p, q) - math.Log(q)
|
|
ib := int(math.Max(xb/alnsml, 0.0))
|
|
term := math.Exp(xb - float64(ib)*alnsml)
|
|
c := 1.0 / (1.0 - y)
|
|
p1 := q * c / (p + q - 1.0)
|
|
finsum := 0.0
|
|
n := int(q)
|
|
if q == float64(n) {
|
|
n = n - 1
|
|
}
|
|
for i := 1; i <= n; i++ {
|
|
if p1 <= 1 && term/eps <= finsum {
|
|
break
|
|
}
|
|
xi := float64(i)
|
|
term = (q - xi + 1.0) * c * term / (p + q - xi)
|
|
if term > 1.0 {
|
|
ib = ib - 1
|
|
term = term * sml
|
|
}
|
|
if ib == 0 {
|
|
finsum = finsum + term
|
|
}
|
|
}
|
|
ans = ans + finsum
|
|
}
|
|
if y != x || p != pin {
|
|
ans = 1.0 - ans
|
|
}
|
|
ans = math.Max(math.Min(ans, 1.0), 0.0)
|
|
return ans
|
|
}
|
|
|
|
// pbeta returns distribution function of the beta distribution.
|
|
func pbeta(x, pin, qin float64) (ans float64) {
|
|
eps := d1mach(3)
|
|
alneps := math.Log(eps)
|
|
sml := d1mach(1)
|
|
alnsml := math.Log(sml)
|
|
y := x
|
|
p := pin
|
|
q := qin
|
|
if p/(p+q) < x {
|
|
y = 1.0 - y
|
|
p = qin
|
|
q = pin
|
|
}
|
|
if (p+q)*y/(p+1.0) < eps {
|
|
xb := p*math.Log(math.Max(y, sml)) - math.Log(p) - logBeta(p, q)
|
|
if xb > alnsml && y != 0.0 {
|
|
ans = math.Exp(xb)
|
|
}
|
|
if y != x || p != pin {
|
|
ans = 1.0 - ans
|
|
}
|
|
} else {
|
|
ps := q - math.Floor(q)
|
|
if ps == 0.0 {
|
|
ps = 1.0
|
|
}
|
|
xb := p*math.Log(y) - logBeta(ps, p) - math.Log(p)
|
|
if xb >= alnsml {
|
|
ans = math.Exp(xb)
|
|
term := ans * p
|
|
if ps != 1.0 {
|
|
n := int(math.Max(alneps/math.Log(y), 4.0))
|
|
for i := 1; i <= n; i++ {
|
|
xi := float64(i)
|
|
term = term * (xi - ps) * y / xi
|
|
ans = ans + term/(p+xi)
|
|
}
|
|
}
|
|
}
|
|
ans = pbetaRaw(alnsml, ans, eps, p, pin, q, sml, x, y)
|
|
}
|
|
return ans
|
|
}
|
|
|
|
// betainvProbIterator is a part of betainv for the inverse of the beta
|
|
// function.
|
|
func betainvProbIterator(alpha1, alpha3, beta1, beta2, beta3, logBeta, maxCumulative, prob1, prob2 float64) float64 {
|
|
var i, j, prev, prop4 float64
|
|
j = 1
|
|
for prob := 0; prob < 1000; prob++ {
|
|
prop3 := pbeta(beta3, alpha1, beta1)
|
|
prop3 = (prop3 - prob1) * math.Exp(logBeta+prob2*math.Log(beta3)+beta2*math.Log(1.0-beta3))
|
|
if prop3*prop4 <= 0 {
|
|
prev = math.Max(math.Abs(j), maxCumulative)
|
|
}
|
|
h := 1.0
|
|
for iteratorCount := 0; iteratorCount < 1000; iteratorCount++ {
|
|
j = h * prop3
|
|
if math.Abs(j) < prev {
|
|
i = beta3 - j
|
|
if i >= 0 && i <= 1.0 {
|
|
if prev <= alpha3 {
|
|
return beta3
|
|
}
|
|
if math.Abs(prop3) <= alpha3 {
|
|
return beta3
|
|
}
|
|
if i != 0 && i != 1.0 {
|
|
break
|
|
}
|
|
}
|
|
}
|
|
h /= 3.0
|
|
}
|
|
if i == beta3 {
|
|
return beta3
|
|
}
|
|
beta3, prop4 = i, prop3
|
|
}
|
|
return beta3
|
|
}
|
|
|
|
// calcBetainv is an implementation for the quantile of the beta
|
|
// distribution.
|
|
func calcBetainv(probability, alpha, beta, lower, upper float64) float64 {
|
|
minCumulative, maxCumulative := 1.0e-300, 3.0e-308
|
|
lowerBound, upperBound := maxCumulative, 1.0-2.22e-16
|
|
needSwap := false
|
|
var alpha1, alpha2, beta1, beta2, beta3, prob1, x, y float64
|
|
if probability <= 0.5 {
|
|
prob1, alpha1, beta1 = probability, alpha, beta
|
|
} else {
|
|
prob1, alpha1, beta1, needSwap = 1.0-probability, beta, alpha, true
|
|
}
|
|
logBetaNum := logBeta(alpha, beta)
|
|
prob2 := math.Sqrt(-math.Log(prob1 * prob1))
|
|
prob3 := prob2 - (prob2*0.27061+2.3075)/(prob2*(prob2*0.04481+0.99229)+1)
|
|
if alpha1 > 1 && beta1 > 1 {
|
|
alpha2, beta2, prob2 = 1/(alpha1+alpha1-1), 1/(beta1+beta1-1), (prob3*prob3-3)/6
|
|
x = 2 / (alpha2 + beta2)
|
|
y = prob3*math.Sqrt(x+prob2)/x - (beta2-alpha2)*(prob2+5/6.0-2/(x*3))
|
|
beta3 = alpha1 / (alpha1 + beta1*math.Exp(y+y))
|
|
} else {
|
|
beta2, prob2 = 1/(beta1*9), beta1+beta1
|
|
beta2 = prob2 * math.Pow(1-beta2+prob3*math.Sqrt(beta2), 3)
|
|
if beta2 <= 0 {
|
|
beta3 = 1 - math.Exp((math.Log((1-prob1)*beta1)+logBetaNum)/beta1)
|
|
} else {
|
|
beta2 = (prob2 + alpha1*4 - 2) / beta2
|
|
if beta2 <= 1 {
|
|
beta3 = math.Exp((logBetaNum + math.Log(alpha1*prob1)) / alpha1)
|
|
} else {
|
|
beta3 = 1 - 2/(beta2+1)
|
|
}
|
|
}
|
|
}
|
|
beta2, prob2 = 1-beta1, 1-alpha1
|
|
if beta3 < lowerBound {
|
|
beta3 = lowerBound
|
|
} else if beta3 > upperBound {
|
|
beta3 = upperBound
|
|
}
|
|
alpha3 := math.Max(minCumulative, math.Pow(10.0, -13.0-2.5/(alpha1*alpha1)-0.5/(prob1*prob1)))
|
|
beta3 = betainvProbIterator(alpha1, alpha3, beta1, beta2, beta3, logBetaNum, maxCumulative, prob1, prob2)
|
|
if needSwap {
|
|
beta3 = 1.0 - beta3
|
|
}
|
|
return (upper-lower)*beta3 + lower
|
|
}
|
|
|
|
// betainv is an implementation of the formula functions BETAINV and
|
|
// BETA.INV.
|
|
func (fn *formulaFuncs) betainv(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 3 arguments", name))
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at most 5 arguments", name))
|
|
}
|
|
probability := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if probability.Number <= 0 || probability.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
alpha := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
beta := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if beta.Type != ArgNumber {
|
|
return beta
|
|
}
|
|
if alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
a, b := newNumberFormulaArg(0), newNumberFormulaArg(1)
|
|
if argsList.Len() > 3 {
|
|
if a = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); a.Type != ArgNumber {
|
|
return a
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if b = argsList.Back().Value.(formulaArg).ToNumber(); b.Type != ArgNumber {
|
|
return b
|
|
}
|
|
}
|
|
if a.Number == b.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(calcBetainv(probability.Number, alpha.Number, beta.Number, a.Number, b.Number))
|
|
}
|
|
|
|
// BETAINV function uses an iterative procedure to calculate the inverse of
|
|
// the cumulative beta probability density function for a supplied
|
|
// probability. The syntax of the function is:
|
|
//
|
|
// BETAINV(probability,alpha,beta,[A],[B])
|
|
func (fn *formulaFuncs) BETAINV(argsList *list.List) formulaArg {
|
|
return fn.betainv("BETAINV", argsList)
|
|
}
|
|
|
|
// BETAdotINV function uses an iterative procedure to calculate the inverse of
|
|
// the cumulative beta probability density function for a supplied
|
|
// probability. The syntax of the function is:
|
|
//
|
|
// BETA.INV(probability,alpha,beta,[A],[B])
|
|
func (fn *formulaFuncs) BETAdotINV(argsList *list.List) formulaArg {
|
|
return fn.betainv("BETA.INV", argsList)
|
|
}
|
|
|
|
// incompleteGamma is an implementation of the incomplete gamma function.
|
|
func incompleteGamma(a, x float64) float64 {
|
|
max := 32
|
|
summer := 0.0
|
|
for n := 0; n <= max; n++ {
|
|
divisor := a
|
|
for i := 1; i <= n; i++ {
|
|
divisor *= a + float64(i)
|
|
}
|
|
summer += math.Pow(x, float64(n)) / divisor
|
|
}
|
|
return math.Pow(x, a) * math.Exp(0-x) * summer
|
|
}
|
|
|
|
// binomCoeff implement binomial coefficient calculation.
|
|
func binomCoeff(n, k float64) float64 {
|
|
return fact(n) / (fact(k) * fact(n-k))
|
|
}
|
|
|
|
// binomdist implement binomial distribution calculation.
|
|
func binomdist(x, n, p float64) float64 {
|
|
return binomCoeff(n, x) * math.Pow(p, x) * math.Pow(1-p, n-x)
|
|
}
|
|
|
|
// BINOMdotDIST function returns the Binomial Distribution probability for a
|
|
// given number of successes from a specified number of trials. The syntax of
|
|
// the function is:
|
|
//
|
|
// BINOM.DIST(number_s,trials,probability_s,cumulative)
|
|
func (fn *formulaFuncs) BINOMdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BINOM.DIST requires 4 arguments")
|
|
}
|
|
return fn.BINOMDIST(argsList)
|
|
}
|
|
|
|
// BINOMDIST function returns the Binomial Distribution probability of a
|
|
// specified number of successes out of a specified number of trials. The
|
|
// syntax of the function is:
|
|
//
|
|
// BINOMDIST(number_s,trials,probability_s,cumulative)
|
|
func (fn *formulaFuncs) BINOMDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BINOMDIST requires 4 arguments")
|
|
}
|
|
var s, trials, probability, cumulative formulaArg
|
|
if s = argsList.Front().Value.(formulaArg).ToNumber(); s.Type != ArgNumber {
|
|
return s
|
|
}
|
|
if trials = argsList.Front().Next().Value.(formulaArg).ToNumber(); trials.Type != ArgNumber {
|
|
return trials
|
|
}
|
|
if s.Number < 0 || s.Number > trials.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if probability = argsList.Back().Prev().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
|
|
if probability.Number < 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if cumulative.Number == 1 {
|
|
bm := 0.0
|
|
for i := 0; i <= int(s.Number); i++ {
|
|
bm += binomdist(float64(i), trials.Number, probability.Number)
|
|
}
|
|
return newNumberFormulaArg(bm)
|
|
}
|
|
return newNumberFormulaArg(binomdist(s.Number, trials.Number, probability.Number))
|
|
}
|
|
|
|
// BINOMdotDISTdotRANGE function returns the Binomial Distribution probability
|
|
// for the number of successes from a specified number of trials falling into
|
|
// a specified range.
|
|
//
|
|
// BINOM.DIST.RANGE(trials,probability_s,number_s,[number_s2])
|
|
func (fn *formulaFuncs) BINOMdotDISTdotRANGE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BINOM.DIST.RANGE requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BINOM.DIST.RANGE requires at most 4 arguments")
|
|
}
|
|
trials := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if trials.Type != ArgNumber {
|
|
return trials
|
|
}
|
|
probability := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if probability.Number < 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
num1 := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if num1.Type != ArgNumber {
|
|
return num1
|
|
}
|
|
if num1.Number < 0 || num1.Number > trials.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
num2 := num1
|
|
if argsList.Len() > 3 {
|
|
if num2 = argsList.Back().Value.(formulaArg).ToNumber(); num2.Type != ArgNumber {
|
|
return num2
|
|
}
|
|
}
|
|
if num2.Number < 0 || num2.Number > trials.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
sum := 0.0
|
|
for i := num1.Number; i <= num2.Number; i++ {
|
|
sum += binomdist(i, trials.Number, probability.Number)
|
|
}
|
|
return newNumberFormulaArg(sum)
|
|
}
|
|
|
|
// binominv implement inverse of the binomial distribution calculation.
|
|
func binominv(n, p, alpha float64) float64 {
|
|
q, i, sum, max := 1-p, 0.0, 0.0, 0.0
|
|
n = math.Floor(n)
|
|
if q > p {
|
|
factor := math.Pow(q, n)
|
|
sum = factor
|
|
for i = 0; i < n && sum < alpha; i++ {
|
|
factor *= (n - i) / (i + 1) * p / q
|
|
sum += factor
|
|
}
|
|
return i
|
|
}
|
|
factor := math.Pow(p, n)
|
|
sum, max = 1-factor, n
|
|
for i = 0; i < max && sum >= alpha; i++ {
|
|
factor *= (n - i) / (i + 1) * q / p
|
|
sum -= factor
|
|
}
|
|
return n - i
|
|
}
|
|
|
|
// BINOMdotINV function returns the inverse of the Cumulative Binomial
|
|
// Distribution. The syntax of the function is:
|
|
//
|
|
// BINOM.INV(trials,probability_s,alpha)
|
|
func (fn *formulaFuncs) BINOMdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "BINOM.INV requires 3 numeric arguments")
|
|
}
|
|
trials := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if trials.Type != ArgNumber {
|
|
return trials
|
|
}
|
|
if trials.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
probability := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if probability.Number <= 0 || probability.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
alpha := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
if alpha.Number <= 0 || alpha.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(binominv(trials.Number, probability.Number, alpha.Number))
|
|
}
|
|
|
|
// CHIDIST function calculates the right-tailed probability of the chi-square
|
|
// distribution. The syntax of the function is:
|
|
//
|
|
// CHIDIST(x,degrees_freedom)
|
|
func (fn *formulaFuncs) CHIDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHIDIST requires 2 numeric arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
degrees := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
logSqrtPi, sqrtPi := math.Log(math.Sqrt(math.Pi)), 1/math.Sqrt(math.Pi)
|
|
var e, s, z, c, y float64
|
|
a, x1, even := x.Number/2, x.Number, int(degrees.Number)%2 == 0
|
|
if degrees.Number > 1 {
|
|
y = math.Exp(-a)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg(-math.Sqrt(x1)))
|
|
o := fn.NORMSDIST(args)
|
|
s = 2 * o.Number
|
|
if even {
|
|
s = y
|
|
}
|
|
if degrees.Number > 2 {
|
|
x1 = (degrees.Number - 1) / 2
|
|
z = 0.5
|
|
if even {
|
|
z = 1
|
|
}
|
|
if a > 20 {
|
|
e = logSqrtPi
|
|
if even {
|
|
e = 0
|
|
}
|
|
c = math.Log(a)
|
|
for z <= x1 {
|
|
e = math.Log(z) + e
|
|
s += math.Exp(c*z - a - e)
|
|
z++
|
|
}
|
|
return newNumberFormulaArg(s)
|
|
}
|
|
e = sqrtPi / math.Sqrt(a)
|
|
if even {
|
|
e = 1
|
|
}
|
|
c = 0
|
|
for z <= x1 {
|
|
e = e * (a / z)
|
|
c = c + e
|
|
z++
|
|
}
|
|
return newNumberFormulaArg(c*y + s)
|
|
}
|
|
return newNumberFormulaArg(s)
|
|
}
|
|
|
|
// CHIINV function calculates the inverse of the right-tailed probability of
|
|
// the Chi-Square Distribution. The syntax of the function is:
|
|
//
|
|
// CHIINV(probability,deg_freedom)
|
|
func (fn *formulaFuncs) CHIINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHIINV requires 2 numeric arguments")
|
|
}
|
|
probability := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if probability.Number <= 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
deg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if deg.Type != ArgNumber {
|
|
return deg
|
|
}
|
|
if deg.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(gammainv(1-probability.Number, 0.5*deg.Number, 2.0))
|
|
}
|
|
|
|
// CHITEST function uses the chi-square test to calculate the probability that
|
|
// the differences between two supplied data sets (of observed and expected
|
|
// frequencies), are likely to be simply due to sampling error, or if they are
|
|
// likely to be real. The syntax of the function is:
|
|
//
|
|
// CHITEST(actual_range,expected_range)
|
|
func (fn *formulaFuncs) CHITEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHITEST requires 2 arguments")
|
|
}
|
|
actual, expected := argsList.Front().Value.(formulaArg), argsList.Back().Value.(formulaArg)
|
|
actualList, expectedList := actual.ToList(), expected.ToList()
|
|
rows := len(actual.Matrix)
|
|
columns := len(actualList) / rows
|
|
if len(actualList) != len(expectedList) || len(actualList) == 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var result float64
|
|
var degrees int
|
|
for i := 0; i < len(actualList); i++ {
|
|
a, e := actualList[i].ToNumber(), expectedList[i].ToNumber()
|
|
if a.Type == ArgNumber && e.Type == ArgNumber {
|
|
if e.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
if e.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
result += (a.Number - e.Number) * (a.Number - e.Number) / e.Number
|
|
}
|
|
}
|
|
if rows == 1 {
|
|
degrees = columns - 1
|
|
} else if columns == 1 {
|
|
degrees = rows - 1
|
|
} else {
|
|
degrees = (columns - 1) * (rows - 1)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg(result))
|
|
args.PushBack(newNumberFormulaArg(float64(degrees)))
|
|
return fn.CHIDIST(args)
|
|
}
|
|
|
|
// getGammaSeries calculates a power-series of the gamma function.
|
|
func getGammaSeries(fA, fX float64) float64 {
|
|
var (
|
|
fHalfMachEps = 2.22045e-016 / 2
|
|
fDenomfactor = fA
|
|
fSummand = 1 / fA
|
|
fSum = fSummand
|
|
nCount = 1
|
|
)
|
|
for fSummand/fSum > fHalfMachEps && nCount <= 10000 {
|
|
fDenomfactor = fDenomfactor + 1
|
|
fSummand = fSummand * fX / fDenomfactor
|
|
fSum = fSum + fSummand
|
|
nCount = nCount + 1
|
|
}
|
|
return fSum
|
|
}
|
|
|
|
// getGammaContFraction returns continued fraction with odd items of the gamma
|
|
// function.
|
|
func getGammaContFraction(fA, fX float64) float64 {
|
|
var (
|
|
fBigInv = 2.22045e-016
|
|
fHalfMachEps = fBigInv / 2
|
|
fBig = 1 / fBigInv
|
|
fCount = 0.0
|
|
fY = 1 - fA
|
|
fDenom = fX + 2 - fA
|
|
fPkm1 = fX + 1
|
|
fPkm2 = 1.0
|
|
fQkm1 = fDenom * fX
|
|
fQkm2 = fX
|
|
fApprox = fPkm1 / fQkm1
|
|
bFinished = false
|
|
)
|
|
for !bFinished && fCount < 10000 {
|
|
fCount = fCount + 1
|
|
fY = fY + 1
|
|
fDenom = fDenom + 2
|
|
var (
|
|
fNum = fY * fCount
|
|
f1 = fPkm1 * fDenom
|
|
f2 = fPkm2 * fNum
|
|
fPk = math.Nextafter(f1, f1) - math.Nextafter(f2, f2)
|
|
f3 = fQkm1 * fDenom
|
|
f4 = fQkm2 * fNum
|
|
fQk = math.Nextafter(f3, f3) - math.Nextafter(f4, f4)
|
|
)
|
|
if fQk != 0 {
|
|
fR := fPk / fQk
|
|
bFinished = math.Abs((fApprox-fR)/fR) <= fHalfMachEps
|
|
fApprox = fR
|
|
}
|
|
fPkm2, fPkm1, fQkm2, fQkm1 = fPkm1, fPk, fQkm1, fQk
|
|
if math.Abs(fPk) > fBig {
|
|
// reduce a fraction does not change the value
|
|
fPkm2 = fPkm2 * fBigInv
|
|
fPkm1 = fPkm1 * fBigInv
|
|
fQkm2 = fQkm2 * fBigInv
|
|
fQkm1 = fQkm1 * fBigInv
|
|
}
|
|
}
|
|
return fApprox
|
|
}
|
|
|
|
// getLogGammaHelper is a part of implementation of the function getLogGamma.
|
|
func getLogGammaHelper(fZ float64) float64 {
|
|
_fg := 6.024680040776729583740234375
|
|
zgHelp := fZ + _fg - 0.5
|
|
return math.Log(getLanczosSum(fZ)) + (fZ-0.5)*math.Log(zgHelp) - zgHelp
|
|
}
|
|
|
|
// getGammaHelper is a part of implementation of the function getLogGamma.
|
|
func getGammaHelper(fZ float64) float64 {
|
|
var (
|
|
gamma = getLanczosSum(fZ)
|
|
fg = 6.024680040776729583740234375
|
|
zgHelp = fZ + fg - 0.5
|
|
// avoid intermediate overflow
|
|
halfpower = math.Pow(zgHelp, fZ/2-0.25)
|
|
)
|
|
gamma *= halfpower
|
|
gamma /= math.Exp(zgHelp)
|
|
gamma *= halfpower
|
|
if fZ <= 20 && fZ == math.Floor(fZ) {
|
|
gamma = math.Round(gamma)
|
|
}
|
|
return gamma
|
|
}
|
|
|
|
// getLogGamma calculates the natural logarithm of the gamma function.
|
|
func getLogGamma(fZ float64) float64 {
|
|
fMaxGammaArgument := 171.624376956302
|
|
if fZ >= fMaxGammaArgument {
|
|
return getLogGammaHelper(fZ)
|
|
}
|
|
if fZ >= 1.0 {
|
|
return math.Log(getGammaHelper(fZ))
|
|
}
|
|
if fZ >= 0.5 {
|
|
return math.Log(getGammaHelper(fZ+1) / fZ)
|
|
}
|
|
return getLogGammaHelper(fZ+2) - math.Log(fZ+1) - math.Log(fZ)
|
|
}
|
|
|
|
// getLowRegIGamma returns lower regularized incomplete gamma function.
|
|
func getLowRegIGamma(fA, fX float64) float64 {
|
|
lnFactor := fA*math.Log(fX) - fX - getLogGamma(fA)
|
|
factor := math.Exp(lnFactor)
|
|
if fX > fA+1 {
|
|
return 1 - factor*getGammaContFraction(fA, fX)
|
|
}
|
|
return factor * getGammaSeries(fA, fX)
|
|
}
|
|
|
|
// getChiSqDistCDF returns left tail for the Chi-Square distribution.
|
|
func getChiSqDistCDF(fX, fDF float64) float64 {
|
|
if fX <= 0 {
|
|
return 0
|
|
}
|
|
return getLowRegIGamma(fDF/2, fX/2)
|
|
}
|
|
|
|
// getChiSqDistPDF calculates the probability density function for the
|
|
// Chi-Square distribution.
|
|
func getChiSqDistPDF(fX, fDF float64) float64 {
|
|
if fDF*fX > 1391000 {
|
|
return math.Exp((0.5*fDF-1)*math.Log(fX*0.5) - 0.5*fX - math.Log(2) - getLogGamma(0.5*fDF))
|
|
}
|
|
var fCount, fValue float64
|
|
if math.Mod(fDF, 2) < 0.5 {
|
|
fValue = 0.5
|
|
fCount = 2
|
|
} else {
|
|
fValue = 1 / math.Sqrt(fX*2*math.Pi)
|
|
fCount = 1
|
|
}
|
|
for fCount < fDF {
|
|
fValue *= fX / fCount
|
|
fCount += 2
|
|
}
|
|
if fX >= 1425 {
|
|
fValue = math.Exp(math.Log(fValue) - fX/2)
|
|
} else {
|
|
fValue *= math.Exp(-fX / 2)
|
|
}
|
|
return fValue
|
|
}
|
|
|
|
// CHISQdotDIST function calculates the Probability Density Function or the
|
|
// Cumulative Distribution Function for the Chi-Square Distribution. The
|
|
// syntax of the function is:
|
|
//
|
|
// CHISQ.DIST(x,degrees_freedom,cumulative)
|
|
func (fn *formulaFuncs) CHISQdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHISQ.DIST requires 3 arguments")
|
|
}
|
|
var x, degrees, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if degrees = argsList.Front().Next().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if x.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
maxDeg := math.Pow10(10)
|
|
if degrees.Number < 1 || degrees.Number >= maxDeg {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(getChiSqDistCDF(x.Number, degrees.Number))
|
|
}
|
|
return newNumberFormulaArg(getChiSqDistPDF(x.Number, degrees.Number))
|
|
}
|
|
|
|
// CHISQdotDISTdotRT function calculates the right-tailed probability of the
|
|
// Chi-Square Distribution. The syntax of the function is:
|
|
//
|
|
// CHISQ.DIST.RT(x,degrees_freedom)
|
|
func (fn *formulaFuncs) CHISQdotDISTdotRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHISQ.DIST.RT requires 2 numeric arguments")
|
|
}
|
|
return fn.CHIDIST(argsList)
|
|
}
|
|
|
|
// CHISQdotTEST function performs the chi-square test on two supplied data sets
|
|
// (of observed and expected frequencies), and returns the probability that
|
|
// the differences between the sets are simply due to sampling error. The
|
|
// syntax of the function is:
|
|
//
|
|
// CHISQ.TEST(actual_range,expected_range)
|
|
func (fn *formulaFuncs) CHISQdotTEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHISQ.TEST requires 2 arguments")
|
|
}
|
|
return fn.CHITEST(argsList)
|
|
}
|
|
|
|
// hasChangeOfSign check if the sign has been changed.
|
|
func hasChangeOfSign(u, w float64) bool {
|
|
return (u < 0 && w > 0) || (u > 0 && w < 0)
|
|
}
|
|
|
|
// calcInverseIterator directly maps the required parameters for inverse
|
|
// distribution functions.
|
|
type calcInverseIterator struct {
|
|
name string
|
|
fp, fDF, nT float64
|
|
}
|
|
|
|
// callBack implements the callback function for the inverse iterator.
|
|
func (iterator *calcInverseIterator) callBack(x float64) float64 {
|
|
if iterator.name == "CHISQ.INV" {
|
|
return iterator.fp - getChiSqDistCDF(x, iterator.fDF)
|
|
}
|
|
return iterator.fp - getTDist(x, iterator.fDF, iterator.nT)
|
|
}
|
|
|
|
// inverseQuadraticInterpolation inverse quadratic interpolation with
|
|
// additional brackets.
|
|
func inverseQuadraticInterpolation(iterator calcInverseIterator, fAx, fAy, fBx, fBy float64) float64 {
|
|
fYEps := 1.0e-307
|
|
fXEps := 2.22045e-016
|
|
fPx, fPy, fQx, fQy, fRx, fRy := fAx, fAy, fBx, fBy, fAx, fAy
|
|
fSx := 0.5 * (fAx + fBx)
|
|
bHasToInterpolate := true
|
|
nCount := 0
|
|
for nCount < 500 && math.Abs(fRy) > fYEps && (fBx-fAx) > math.Max(math.Abs(fAx), math.Abs(fBx))*fXEps {
|
|
if bHasToInterpolate {
|
|
if fPy != fQy && fQy != fRy && fRy != fPy {
|
|
fSx = fPx*fRy*fQy/(fRy-fPy)/(fQy-fPy) + fRx*fQy*fPy/(fQy-fRy)/(fPy-fRy) +
|
|
fQx*fPy*fRy/(fPy-fQy)/(fRy-fQy)
|
|
bHasToInterpolate = (fAx < fSx) && (fSx < fBx)
|
|
} else {
|
|
bHasToInterpolate = false
|
|
}
|
|
}
|
|
if !bHasToInterpolate {
|
|
fSx = 0.5 * (fAx + fBx)
|
|
fQx, fQy = fBx, fBy
|
|
bHasToInterpolate = true
|
|
}
|
|
fPx, fQx, fRx, fPy, fQy = fQx, fRx, fSx, fQy, fRy
|
|
fRy = iterator.callBack(fSx)
|
|
if hasChangeOfSign(fAy, fRy) {
|
|
fBx, fBy = fRx, fRy
|
|
} else {
|
|
fAx, fAy = fRx, fRy
|
|
}
|
|
bHasToInterpolate = bHasToInterpolate && (math.Abs(fRy)*2 <= math.Abs(fQy))
|
|
nCount++
|
|
}
|
|
return fRx
|
|
}
|
|
|
|
// calcIterateInverse function calculates the iteration for inverse
|
|
// distributions.
|
|
func calcIterateInverse(iterator calcInverseIterator, fAx, fBx float64) float64 {
|
|
fAy, fBy := iterator.callBack(fAx), iterator.callBack(fBx)
|
|
var fTemp float64
|
|
var nCount int
|
|
for nCount = 0; nCount < 1000 && !hasChangeOfSign(fAy, fBy); nCount++ {
|
|
if math.Abs(fAy) <= math.Abs(fBy) {
|
|
fTemp = fAx
|
|
fAx += 2 * (fAx - fBx)
|
|
if fAx < 0 {
|
|
fAx = 0
|
|
}
|
|
fBx = fTemp
|
|
fBy = fAy
|
|
fAy = iterator.callBack(fAx)
|
|
} else {
|
|
fTemp = fBx
|
|
fBx += 2 * (fBx - fAx)
|
|
fAx = fTemp
|
|
fAy = fBy
|
|
fBy = iterator.callBack(fBx)
|
|
}
|
|
}
|
|
if fAy == 0 || fBy == 0 {
|
|
return 0
|
|
}
|
|
return inverseQuadraticInterpolation(iterator, fAx, fAy, fBx, fBy)
|
|
}
|
|
|
|
// CHISQdotINV function calculates the inverse of the left-tailed probability
|
|
// of the Chi-Square Distribution. The syntax of the function is:
|
|
//
|
|
// CHISQ.INV(probability,degrees_freedom)
|
|
func (fn *formulaFuncs) CHISQdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHISQ.INV requires 2 numeric arguments")
|
|
}
|
|
var probability, degrees formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if degrees = argsList.Back().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if probability.Number < 0 || probability.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if degrees.Number < 1 || degrees.Number > math.Pow10(10) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(calcIterateInverse(calcInverseIterator{
|
|
name: "CHISQ.INV",
|
|
fp: probability.Number,
|
|
fDF: degrees.Number,
|
|
}, degrees.Number/2, degrees.Number))
|
|
}
|
|
|
|
// CHISQdotINVdotRT function calculates the inverse of the right-tailed
|
|
// probability of the Chi-Square Distribution. The syntax of the function is:
|
|
//
|
|
// CHISQ.INV.RT(probability,degrees_freedom)
|
|
func (fn *formulaFuncs) CHISQdotINVdotRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHISQ.INV.RT requires 2 numeric arguments")
|
|
}
|
|
return fn.CHIINV(argsList)
|
|
}
|
|
|
|
// confidence is an implementation of the formula functions CONFIDENCE and
|
|
// CONFIDENCE.NORM.
|
|
func (fn *formulaFuncs) confidence(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 3 numeric arguments", name))
|
|
}
|
|
alpha := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
if alpha.Number <= 0 || alpha.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
stdDev := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if stdDev.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
size := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if size.Type != ArgNumber {
|
|
return size
|
|
}
|
|
if size.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New()
|
|
args.Init()
|
|
args.PushBack(newNumberFormulaArg(alpha.Number / 2))
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
return newNumberFormulaArg(-fn.NORMINV(args).Number * (stdDev.Number / math.Sqrt(size.Number)))
|
|
}
|
|
|
|
// CONFIDENCE function uses a Normal Distribution to calculate a confidence
|
|
// value that can be used to construct the Confidence Interval for a
|
|
// population mean, for a supplied probability and sample size. It is assumed
|
|
// that the standard deviation of the population is known. The syntax of the
|
|
// function is:
|
|
//
|
|
// CONFIDENCE(alpha,standard_dev,size)
|
|
func (fn *formulaFuncs) CONFIDENCE(argsList *list.List) formulaArg {
|
|
return fn.confidence("CONFIDENCE", argsList)
|
|
}
|
|
|
|
// CONFIDENCEdotNORM function uses a Normal Distribution to calculate a
|
|
// confidence value that can be used to construct the confidence interval for
|
|
// a population mean, for a supplied probability and sample size. It is
|
|
// assumed that the standard deviation of the population is known. The syntax
|
|
// of the function is:
|
|
//
|
|
// CONFIDENCE.NORM(alpha,standard_dev,size)
|
|
func (fn *formulaFuncs) CONFIDENCEdotNORM(argsList *list.List) formulaArg {
|
|
return fn.confidence("CONFIDENCE.NORM", argsList)
|
|
}
|
|
|
|
// CONFIDENCEdotT function uses a Student's T-Distribution to calculate a
|
|
// confidence value that can be used to construct the confidence interval for
|
|
// a population mean, for a supplied probablity and supplied sample size. It
|
|
// is assumed that the standard deviation of the population is known. The
|
|
// syntax of the function is:
|
|
//
|
|
// CONFIDENCE.T(alpha,standard_dev,size)
|
|
func (fn *formulaFuncs) CONFIDENCEdotT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CONFIDENCE.T requires 3 arguments")
|
|
}
|
|
var alpha, standardDev, size formulaArg
|
|
if alpha = argsList.Front().Value.(formulaArg).ToNumber(); alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
if standardDev = argsList.Front().Next().Value.(formulaArg).ToNumber(); standardDev.Type != ArgNumber {
|
|
return standardDev
|
|
}
|
|
if size = argsList.Back().Value.(formulaArg).ToNumber(); size.Type != ArgNumber {
|
|
return size
|
|
}
|
|
if alpha.Number <= 0 || alpha.Number >= 1 || standardDev.Number <= 0 || size.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if size.Number == 1 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(standardDev.Number * calcIterateInverse(calcInverseIterator{
|
|
name: "CONFIDENCE.T",
|
|
fp: alpha.Number,
|
|
fDF: size.Number - 1,
|
|
nT: 2,
|
|
}, size.Number/2, size.Number) / math.Sqrt(size.Number))
|
|
}
|
|
|
|
// covar is an implementation of the formula functions COVAR, COVARIANCE.P and
|
|
// COVARIANCE.S.
|
|
func (fn *formulaFuncs) covar(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 arguments", name))
|
|
}
|
|
array1 := argsList.Front().Value.(formulaArg)
|
|
array2 := argsList.Back().Value.(formulaArg)
|
|
left, right := array1.ToList(), array2.ToList()
|
|
n := len(left)
|
|
if n != len(right) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
l1, l2 := list.New(), list.New()
|
|
l1.PushBack(array1)
|
|
l2.PushBack(array2)
|
|
result, skip := 0.0, 0
|
|
mean1, mean2 := fn.AVERAGE(l1), fn.AVERAGE(l2)
|
|
for i := 0; i < n; i++ {
|
|
arg1 := left[i].ToNumber()
|
|
arg2 := right[i].ToNumber()
|
|
if arg1.Type == ArgError || arg2.Type == ArgError {
|
|
skip++
|
|
continue
|
|
}
|
|
result += (arg1.Number - mean1.Number) * (arg2.Number - mean2.Number)
|
|
}
|
|
if name == "COVARIANCE.S" {
|
|
return newNumberFormulaArg(result / float64(n-skip-1))
|
|
}
|
|
return newNumberFormulaArg(result / float64(n-skip))
|
|
}
|
|
|
|
// COVAR function calculates the covariance of two supplied sets of values. The
|
|
// syntax of the function is:
|
|
//
|
|
// COVAR(array1,array2)
|
|
func (fn *formulaFuncs) COVAR(argsList *list.List) formulaArg {
|
|
return fn.covar("COVAR", argsList)
|
|
}
|
|
|
|
// COVARIANCEdotP function calculates the population covariance of two supplied
|
|
// sets of values. The syntax of the function is:
|
|
//
|
|
// COVARIANCE.P(array1,array2)
|
|
func (fn *formulaFuncs) COVARIANCEdotP(argsList *list.List) formulaArg {
|
|
return fn.covar("COVARIANCE.P", argsList)
|
|
}
|
|
|
|
// COVARIANCEdotS function calculates the sample covariance of two supplied
|
|
// sets of values. The syntax of the function is:
|
|
//
|
|
// COVARIANCE.S(array1,array2)
|
|
func (fn *formulaFuncs) COVARIANCEdotS(argsList *list.List) formulaArg {
|
|
return fn.covar("COVARIANCE.S", argsList)
|
|
}
|
|
|
|
// calcStringCountSum is part of the implementation countSum.
|
|
func calcStringCountSum(countText bool, count, sum float64, num, arg formulaArg) (float64, float64) {
|
|
if countText && num.Type == ArgError && arg.String != "" {
|
|
count++
|
|
}
|
|
if num.Type == ArgNumber {
|
|
sum += num.Number
|
|
count++
|
|
}
|
|
return count, sum
|
|
}
|
|
|
|
// countSum get count and sum for a formula arguments array.
|
|
func (fn *formulaFuncs) countSum(countText bool, args []formulaArg) (count, sum float64) {
|
|
for _, arg := range args {
|
|
switch arg.Type {
|
|
case ArgNumber:
|
|
if countText || !arg.Boolean {
|
|
sum += arg.Number
|
|
count++
|
|
}
|
|
case ArgString:
|
|
if !countText && (arg.Value() == "TRUE" || arg.Value() == "FALSE") {
|
|
continue
|
|
} else if countText && (arg.Value() == "TRUE" || arg.Value() == "FALSE") {
|
|
num := arg.ToBool()
|
|
if num.Type == ArgNumber {
|
|
count++
|
|
sum += num.Number
|
|
continue
|
|
}
|
|
}
|
|
num := arg.ToNumber()
|
|
count, sum = calcStringCountSum(countText, count, sum, num, arg)
|
|
case ArgList, ArgMatrix:
|
|
cnt, summary := fn.countSum(countText, arg.ToList())
|
|
sum += summary
|
|
count += cnt
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// CORREL function calculates the Pearson Product-Moment Correlation
|
|
// Coefficient for two sets of values. The syntax of the function is:
|
|
//
|
|
// CORREL(array1,array2)
|
|
func (fn *formulaFuncs) CORREL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CORREL requires 2 arguments")
|
|
}
|
|
array1 := argsList.Front().Value.(formulaArg)
|
|
array2 := argsList.Back().Value.(formulaArg)
|
|
left, right := array1.ToList(), array2.ToList()
|
|
n := len(left)
|
|
if n != len(right) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
l1, l2, l3 := list.New(), list.New(), list.New()
|
|
for i := 0; i < n; i++ {
|
|
if lhs, rhs := left[i].ToNumber(), right[i].ToNumber(); lhs.Number != 0 && rhs.Number != 0 {
|
|
l1.PushBack(lhs)
|
|
l2.PushBack(rhs)
|
|
}
|
|
}
|
|
stdev1, stdev2 := fn.STDEV(l1), fn.STDEV(l2)
|
|
if stdev1.Number == 0 || stdev2.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
mean1, mean2, skip := fn.AVERAGE(l1), fn.AVERAGE(l2), 0
|
|
for i := 0; i < n; i++ {
|
|
lhs, rhs := left[i].ToNumber(), right[i].ToNumber()
|
|
if lhs.Number == 0 || rhs.Number == 0 {
|
|
skip++
|
|
continue
|
|
}
|
|
l3.PushBack(newNumberFormulaArg((lhs.Number - mean1.Number) * (rhs.Number - mean2.Number)))
|
|
}
|
|
return newNumberFormulaArg(fn.SUM(l3).Number / float64(n-skip-1) / stdev1.Number / stdev2.Number)
|
|
}
|
|
|
|
// COUNT function returns the count of numeric values in a supplied set of
|
|
// cells or values. This count includes both numbers and dates. The syntax of
|
|
// the function is:
|
|
//
|
|
// COUNT(value1,[value2],...)
|
|
func (fn *formulaFuncs) COUNT(argsList *list.List) formulaArg {
|
|
var count int
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
if num := arg.ToNumber(); num.Type == ArgNumber {
|
|
count++
|
|
}
|
|
case ArgNumber:
|
|
count++
|
|
case ArgMatrix:
|
|
for _, row := range arg.Matrix {
|
|
for _, cell := range row {
|
|
if cell.Type == ArgNumber {
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(float64(count))
|
|
}
|
|
|
|
// COUNTA function returns the number of non-blanks within a supplied set of
|
|
// cells or values. The syntax of the function is:
|
|
//
|
|
// COUNTA(value1,[value2],...)
|
|
func (fn *formulaFuncs) COUNTA(argsList *list.List) formulaArg {
|
|
var count int
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
if arg.String != "" {
|
|
count++
|
|
}
|
|
case ArgNumber:
|
|
count++
|
|
case ArgMatrix:
|
|
for _, row := range arg.ToList() {
|
|
switch row.Type {
|
|
case ArgString:
|
|
if row.String != "" {
|
|
count++
|
|
}
|
|
case ArgNumber:
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(float64(count))
|
|
}
|
|
|
|
// COUNTBLANK function returns the number of blank cells in a supplied range.
|
|
// The syntax of the function is:
|
|
//
|
|
// COUNTBLANK(range)
|
|
func (fn *formulaFuncs) COUNTBLANK(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COUNTBLANK requires 1 argument")
|
|
}
|
|
var count float64
|
|
for _, cell := range argsList.Front().Value.(formulaArg).ToList() {
|
|
if cell.Type == ArgEmpty {
|
|
count++
|
|
}
|
|
}
|
|
return newNumberFormulaArg(count)
|
|
}
|
|
|
|
// COUNTIF function returns the number of cells within a supplied range, that
|
|
// satisfy a given criteria. The syntax of the function is:
|
|
//
|
|
// COUNTIF(range,criteria)
|
|
func (fn *formulaFuncs) COUNTIF(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COUNTIF requires 2 arguments")
|
|
}
|
|
var (
|
|
criteria = formulaCriteriaParser(argsList.Front().Next().Value.(formulaArg).String)
|
|
count float64
|
|
)
|
|
for _, cell := range argsList.Front().Value.(formulaArg).ToList() {
|
|
if ok, _ := formulaCriteriaEval(cell.Value(), criteria); ok {
|
|
count++
|
|
}
|
|
}
|
|
return newNumberFormulaArg(count)
|
|
}
|
|
|
|
// formulaIfsMatch function returns cells reference array which match criteria.
|
|
func formulaIfsMatch(args []formulaArg) (cellRefs []cellRef) {
|
|
for i := 0; i < len(args)-1; i += 2 {
|
|
var match []cellRef
|
|
matrix, criteria := args[i].Matrix, formulaCriteriaParser(args[i+1].Value())
|
|
if i == 0 {
|
|
for rowIdx, row := range matrix {
|
|
for colIdx, col := range row {
|
|
if ok, _ := formulaCriteriaEval(col.Value(), criteria); ok {
|
|
match = append(match, cellRef{Col: colIdx, Row: rowIdx})
|
|
}
|
|
}
|
|
}
|
|
} else {
|
|
for _, ref := range cellRefs {
|
|
value := matrix[ref.Row][ref.Col]
|
|
if ok, _ := formulaCriteriaEval(value.Value(), criteria); ok {
|
|
match = append(match, ref)
|
|
}
|
|
}
|
|
}
|
|
if len(match) == 0 {
|
|
return
|
|
}
|
|
cellRefs = match[:]
|
|
}
|
|
return
|
|
}
|
|
|
|
// COUNTIFS function returns the number of rows within a table, that satisfy a
|
|
// set of given criteria. The syntax of the function is:
|
|
//
|
|
// COUNTIFS(criteria_range1,criteria1,[criteria_range2,criteria2],...)
|
|
func (fn *formulaFuncs) COUNTIFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COUNTIFS requires at least 2 arguments")
|
|
}
|
|
if argsList.Len()%2 != 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var args []formulaArg
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
return newNumberFormulaArg(float64(len(formulaIfsMatch(args))))
|
|
}
|
|
|
|
// CRITBINOM function returns the inverse of the Cumulative Binomial
|
|
// Distribution. I.e. for a specific number of independent trials, the
|
|
// function returns the smallest value (number of successes) for which the
|
|
// cumulative binomial distribution is greater than or equal to a specified
|
|
// value. The syntax of the function is:
|
|
//
|
|
// CRITBINOM(trials,probability_s,alpha)
|
|
func (fn *formulaFuncs) CRITBINOM(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CRITBINOM requires 3 numeric arguments")
|
|
}
|
|
return fn.BINOMdotINV(argsList)
|
|
}
|
|
|
|
// DEVSQ function calculates the sum of the squared deviations from the sample
|
|
// mean. The syntax of the function is:
|
|
//
|
|
// DEVSQ(number1,[number2],...)
|
|
func (fn *formulaFuncs) DEVSQ(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DEVSQ requires at least 1 numeric argument")
|
|
}
|
|
avg, count, result := fn.AVERAGE(argsList), -1, 0.0
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
for _, cell := range arg.Value.(formulaArg).ToList() {
|
|
if cell.Type != ArgNumber {
|
|
continue
|
|
}
|
|
count++
|
|
if count == 0 {
|
|
result = math.Pow(cell.Number-avg.Number, 2)
|
|
continue
|
|
}
|
|
result += math.Pow(cell.Number-avg.Number, 2)
|
|
}
|
|
}
|
|
if count == -1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// FISHER function calculates the Fisher Transformation for a supplied value.
|
|
// The syntax of the function is:
|
|
//
|
|
// FISHER(x)
|
|
func (fn *formulaFuncs) FISHER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FISHER requires 1 numeric argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
arg := token.ToNumber()
|
|
if arg.Type == ArgNumber {
|
|
if arg.Number <= -1 || arg.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(0.5 * math.Log((1+arg.Number)/(1-arg.Number)))
|
|
}
|
|
case ArgNumber:
|
|
if token.Number <= -1 || token.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(0.5 * math.Log((1+token.Number)/(1-token.Number)))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FISHER requires 1 numeric argument")
|
|
}
|
|
|
|
// FISHERINV function calculates the inverse of the Fisher Transformation and
|
|
// returns a value between -1 and +1. The syntax of the function is:
|
|
//
|
|
// FISHERINV(y)
|
|
func (fn *formulaFuncs) FISHERINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FISHERINV requires 1 numeric argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
arg := token.ToNumber()
|
|
if arg.Type == ArgNumber {
|
|
return newNumberFormulaArg((math.Exp(2*arg.Number) - 1) / (math.Exp(2*arg.Number) + 1))
|
|
}
|
|
case ArgNumber:
|
|
return newNumberFormulaArg((math.Exp(2*token.Number) - 1) / (math.Exp(2*token.Number) + 1))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FISHERINV requires 1 numeric argument")
|
|
}
|
|
|
|
// GAMMA function returns the value of the Gamma Function, Γ(n), for a
|
|
// specified number, n. The syntax of the function is:
|
|
//
|
|
// GAMMA(number)
|
|
func (fn *formulaFuncs) GAMMA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMA requires 1 numeric argument")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMA requires 1 numeric argument")
|
|
}
|
|
if number.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(math.Gamma(number.Number))
|
|
}
|
|
|
|
// GAMMAdotDIST function returns the Gamma Distribution, which is frequently
|
|
// used to provide probabilities for values that may have a skewed
|
|
// distribution, such as queuing analysis.
|
|
//
|
|
// GAMMA.DIST(x,alpha,beta,cumulative)
|
|
func (fn *formulaFuncs) GAMMAdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMA.DIST requires 4 arguments")
|
|
}
|
|
return fn.GAMMADIST(argsList)
|
|
}
|
|
|
|
// GAMMADIST function returns the Gamma Distribution, which is frequently used
|
|
// to provide probabilities for values that may have a skewed distribution,
|
|
// such as queuing analysis.
|
|
//
|
|
// GAMMADIST(x,alpha,beta,cumulative)
|
|
func (fn *formulaFuncs) GAMMADIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMADIST requires 4 arguments")
|
|
}
|
|
var x, alpha, beta, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if x.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if alpha = argsList.Front().Next().Value.(formulaArg).ToNumber(); alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
if beta = argsList.Back().Prev().Value.(formulaArg).ToNumber(); beta.Type != ArgNumber {
|
|
return beta
|
|
}
|
|
if alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(incompleteGamma(alpha.Number, x.Number/beta.Number) / math.Gamma(alpha.Number))
|
|
}
|
|
return newNumberFormulaArg((1 / (math.Pow(beta.Number, alpha.Number) * math.Gamma(alpha.Number))) * math.Pow(x.Number, alpha.Number-1) * math.Exp(0-(x.Number/beta.Number)))
|
|
}
|
|
|
|
// gammainv returns the inverse of the Gamma distribution for the specified
|
|
// value.
|
|
func gammainv(probability, alpha, beta float64) float64 {
|
|
xLo, xHi := 0.0, alpha*beta*5
|
|
dx, x, xNew, result := 1024.0, 1.0, 1.0, 0.0
|
|
for i := 0; math.Abs(dx) > 8.88e-016 && i <= 256; i++ {
|
|
result = incompleteGamma(alpha, x/beta) / math.Gamma(alpha)
|
|
e := result - probability
|
|
if e == 0 {
|
|
dx = 0
|
|
} else if e < 0 {
|
|
xLo = x
|
|
} else {
|
|
xHi = x
|
|
}
|
|
pdf := (1 / (math.Pow(beta, alpha) * math.Gamma(alpha))) * math.Pow(x, alpha-1) * math.Exp(0-(x/beta))
|
|
if pdf != 0 {
|
|
dx = e / pdf
|
|
xNew = x - dx
|
|
}
|
|
if xNew < xLo || xNew > xHi || pdf == 0 {
|
|
xNew = (xLo + xHi) / 2
|
|
dx = xNew - x
|
|
}
|
|
x = xNew
|
|
}
|
|
return x
|
|
}
|
|
|
|
// GAMMAdotINV function returns the inverse of the Gamma Cumulative
|
|
// Distribution. The syntax of the function is:
|
|
//
|
|
// GAMMA.INV(probability,alpha,beta)
|
|
func (fn *formulaFuncs) GAMMAdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMA.INV requires 3 arguments")
|
|
}
|
|
return fn.GAMMAINV(argsList)
|
|
}
|
|
|
|
// GAMMAINV function returns the inverse of the Gamma Cumulative Distribution.
|
|
// The syntax of the function is:
|
|
//
|
|
// GAMMAINV(probability,alpha,beta)
|
|
func (fn *formulaFuncs) GAMMAINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMAINV requires 3 arguments")
|
|
}
|
|
var probability, alpha, beta formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if probability.Number < 0 || probability.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if alpha = argsList.Front().Next().Value.(formulaArg).ToNumber(); alpha.Type != ArgNumber {
|
|
return alpha
|
|
}
|
|
if beta = argsList.Back().Value.(formulaArg).ToNumber(); beta.Type != ArgNumber {
|
|
return beta
|
|
}
|
|
if alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(gammainv(probability.Number, alpha.Number, beta.Number))
|
|
}
|
|
|
|
// GAMMALN function returns the natural logarithm of the Gamma Function, Γ
|
|
// (n). The syntax of the function is:
|
|
//
|
|
// GAMMALN(x)
|
|
func (fn *formulaFuncs) GAMMALN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMALN requires 1 numeric argument")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMALN requires 1 numeric argument")
|
|
}
|
|
if x.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(math.Log(math.Gamma(x.Number)))
|
|
}
|
|
|
|
// GAMMALNdotPRECISE function returns the natural logarithm of the Gamma
|
|
// Function, Γ(n). The syntax of the function is:
|
|
//
|
|
// GAMMALN.PRECISE(x)
|
|
func (fn *formulaFuncs) GAMMALNdotPRECISE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAMMALN.PRECISE requires 1 numeric argument")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if x.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(getLogGamma(x.Number))
|
|
}
|
|
|
|
// GAUSS function returns the probability that a member of a standard normal
|
|
// population will fall between the mean and a specified number of standard
|
|
// deviations from the mean. The syntax of the function is:
|
|
//
|
|
// GAUSS(z)
|
|
func (fn *formulaFuncs) GAUSS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GAUSS requires 1 numeric argument")
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 0})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1})
|
|
args.PushBack(newBoolFormulaArg(true))
|
|
normdist := fn.NORMDIST(args)
|
|
if normdist.Type != ArgNumber {
|
|
return normdist
|
|
}
|
|
return newNumberFormulaArg(normdist.Number - 0.5)
|
|
}
|
|
|
|
// GEOMEAN function calculates the geometric mean of a supplied set of values.
|
|
// The syntax of the function is:
|
|
//
|
|
// GEOMEAN(number1,[number2],...)
|
|
func (fn *formulaFuncs) GEOMEAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "GEOMEAN requires at least 1 numeric argument")
|
|
}
|
|
product := fn.PRODUCT(argsList)
|
|
if product.Type != ArgNumber {
|
|
return product
|
|
}
|
|
count := fn.COUNT(argsList)
|
|
min := fn.MIN(argsList)
|
|
if product.Number > 0 && min.Number > 0 {
|
|
return newNumberFormulaArg(math.Pow(product.Number, 1/count.Number))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
|
|
// getNewMatrix create matrix by given columns and rows.
|
|
func getNewMatrix(c, r int) (matrix [][]float64) {
|
|
for i := 0; i < c; i++ {
|
|
for j := 0; j < r; j++ {
|
|
for x := len(matrix); x <= i; x++ {
|
|
matrix = append(matrix, []float64{})
|
|
}
|
|
for y := len(matrix[i]); y <= j; y++ {
|
|
matrix[i] = append(matrix[i], 0)
|
|
}
|
|
matrix[i][j] = 0
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// approxSub subtract two values, if signs are identical and the values are
|
|
// equal, will be returns 0 instead of calculating the subtraction.
|
|
func approxSub(a, b float64) float64 {
|
|
if ((a < 0 && b < 0) || (a > 0 && b > 0)) && math.Abs(a-b) < 2.22045e-016 {
|
|
return 0
|
|
}
|
|
return a - b
|
|
}
|
|
|
|
// matrixClone return a copy of all elements of the original matrix.
|
|
func matrixClone(matrix [][]float64) (cloneMatrix [][]float64) {
|
|
for i := 0; i < len(matrix); i++ {
|
|
for j := 0; j < len(matrix[i]); j++ {
|
|
for x := len(cloneMatrix); x <= i; x++ {
|
|
cloneMatrix = append(cloneMatrix, []float64{})
|
|
}
|
|
for k := len(cloneMatrix[i]); k <= j; k++ {
|
|
cloneMatrix[i] = append(cloneMatrix[i], 0)
|
|
}
|
|
cloneMatrix[i][j] = matrix[i][j]
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// trendGrowthMatrixInfo defined matrix checking result.
|
|
type trendGrowthMatrixInfo struct {
|
|
trendType, nCX, nCY, nRX, nRY, M, N int
|
|
mtxX, mtxY [][]float64
|
|
}
|
|
|
|
// prepareTrendGrowthMtxX is a part of implementation of the trend growth prepare.
|
|
func prepareTrendGrowthMtxX(mtxX [][]float64) [][]float64 {
|
|
var mtx [][]float64
|
|
for i := 0; i < len(mtxX); i++ {
|
|
for j := 0; j < len(mtxX[i]); j++ {
|
|
if mtxX[i][j] == 0 {
|
|
return nil
|
|
}
|
|
for x := len(mtx); x <= j; x++ {
|
|
mtx = append(mtx, []float64{})
|
|
}
|
|
for y := len(mtx[j]); y <= i; y++ {
|
|
mtx[j] = append(mtx[j], 0)
|
|
}
|
|
mtx[j][i] = mtxX[i][j]
|
|
}
|
|
}
|
|
return mtx
|
|
}
|
|
|
|
// prepareTrendGrowthMtxY is a part of implementation of the trend growth prepare.
|
|
func prepareTrendGrowthMtxY(bLOG bool, mtxY [][]float64) [][]float64 {
|
|
var mtx [][]float64
|
|
for i := 0; i < len(mtxY); i++ {
|
|
for j := 0; j < len(mtxY[i]); j++ {
|
|
if mtxY[i][j] == 0 {
|
|
return nil
|
|
}
|
|
for x := len(mtx); x <= j; x++ {
|
|
mtx = append(mtx, []float64{})
|
|
}
|
|
for y := len(mtx[j]); y <= i; y++ {
|
|
mtx[j] = append(mtx[j], 0)
|
|
}
|
|
mtx[j][i] = mtxY[i][j]
|
|
}
|
|
}
|
|
if bLOG {
|
|
var pNewY [][]float64
|
|
for i := 0; i < len(mtxY); i++ {
|
|
for j := 0; j < len(mtxY[i]); j++ {
|
|
fVal := mtxY[i][j]
|
|
if fVal <= 0 {
|
|
return nil
|
|
}
|
|
for x := len(pNewY); x <= j; x++ {
|
|
pNewY = append(pNewY, []float64{})
|
|
}
|
|
for y := len(pNewY[j]); y <= i; y++ {
|
|
pNewY[j] = append(pNewY[j], 0)
|
|
}
|
|
pNewY[j][i] = math.Log(fVal)
|
|
}
|
|
}
|
|
mtx = pNewY
|
|
}
|
|
return mtx
|
|
}
|
|
|
|
// prepareTrendGrowth check and return the result.
|
|
func prepareTrendGrowth(bLOG bool, mtxX, mtxY [][]float64) (*trendGrowthMatrixInfo, formulaArg) {
|
|
var nCX, nRX, M, N, trendType int
|
|
nRY, nCY := len(mtxY), len(mtxY[0])
|
|
cntY := nCY * nRY
|
|
newY := prepareTrendGrowthMtxY(bLOG, mtxY)
|
|
if newY == nil {
|
|
return nil, newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
var newX [][]float64
|
|
if len(mtxX) != 0 {
|
|
nRX, nCX = len(mtxX), len(mtxX[0])
|
|
if newX = prepareTrendGrowthMtxX(mtxX); newX == nil {
|
|
return nil, newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if nCX == nCY && nRX == nRY {
|
|
trendType, M, N = 1, 1, cntY // simple regression
|
|
} else if nCY != 1 && nRY != 1 {
|
|
return nil, newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
} else if nCY == 1 {
|
|
if nRX != nRY {
|
|
return nil, newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
}
|
|
trendType, M, N = 2, nCX, nRY
|
|
} else if nCX != nCY {
|
|
return nil, newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
} else {
|
|
trendType, M, N = 3, nRX, nCY
|
|
}
|
|
} else {
|
|
newX = getNewMatrix(nCY, nRY)
|
|
nCX, nRX = nCY, nRY
|
|
num := 1.0
|
|
for i := 0; i < nRY; i++ {
|
|
for j := 0; j < nCY; j++ {
|
|
newX[j][i] = num
|
|
num++
|
|
}
|
|
}
|
|
trendType, M, N = 1, 1, cntY
|
|
}
|
|
return &trendGrowthMatrixInfo{
|
|
trendType: trendType,
|
|
nCX: nCX,
|
|
nCY: nCY,
|
|
nRX: nRX,
|
|
nRY: nRY,
|
|
M: M,
|
|
N: N,
|
|
mtxX: newX,
|
|
mtxY: newY,
|
|
}, newEmptyFormulaArg()
|
|
}
|
|
|
|
// calcPosition calculate position for matrix by given index.
|
|
func calcPosition(mtx [][]float64, idx int) (row, col int) {
|
|
rowSize := len(mtx[0])
|
|
col = idx
|
|
if rowSize > 1 {
|
|
col = idx / rowSize
|
|
}
|
|
row = idx - col*rowSize
|
|
return
|
|
}
|
|
|
|
// getDouble returns float64 data type value in the matrix by given index.
|
|
func getDouble(mtx [][]float64, idx int) float64 {
|
|
row, col := calcPosition(mtx, idx)
|
|
return mtx[col][row]
|
|
}
|
|
|
|
// putDouble set a float64 data type value in the matrix by given index.
|
|
func putDouble(mtx [][]float64, idx int, val float64) {
|
|
row, col := calcPosition(mtx, idx)
|
|
mtx[col][row] = val
|
|
}
|
|
|
|
// calcMeanOverAll returns mean of the given matrix by over all element.
|
|
func calcMeanOverAll(mtx [][]float64, n int) float64 {
|
|
var sum float64
|
|
for i := 0; i < len(mtx); i++ {
|
|
for j := 0; j < len(mtx[i]); j++ {
|
|
sum += mtx[i][j]
|
|
}
|
|
}
|
|
return sum / float64(n)
|
|
}
|
|
|
|
// calcSumProduct returns uses the matrices as vectors of length M over all
|
|
// element.
|
|
func calcSumProduct(mtxA, mtxB [][]float64, m int) float64 {
|
|
sum := 0.0
|
|
for i := 0; i < m; i++ {
|
|
sum += getDouble(mtxA, i) * getDouble(mtxB, i)
|
|
}
|
|
return sum
|
|
}
|
|
|
|
// calcColumnMeans calculates means of the columns of matrix.
|
|
func calcColumnMeans(mtxX, mtxRes [][]float64, c, r int) {
|
|
for i := 0; i < c; i++ {
|
|
var sum float64
|
|
for k := 0; k < r; k++ {
|
|
sum += mtxX[i][k]
|
|
}
|
|
putDouble(mtxRes, i, sum/float64(r))
|
|
}
|
|
}
|
|
|
|
// calcColumnsDelta calculates subtract of the columns of matrix.
|
|
func calcColumnsDelta(mtx, columnMeans [][]float64, c, r int) {
|
|
for i := 0; i < c; i++ {
|
|
for k := 0; k < r; k++ {
|
|
mtx[i][k] = approxSub(mtx[i][k], getDouble(columnMeans, i))
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcSign returns sign by given value, no mathematical signum, but used to
|
|
// switch between adding and subtracting.
|
|
func calcSign(val float64) float64 {
|
|
if val > 0 {
|
|
return 1
|
|
}
|
|
return -1
|
|
}
|
|
|
|
// calcColsMaximumNorm is a special version for use within QR
|
|
// decomposition. Maximum norm of column index c starting in row index r;
|
|
// matrix A has count n rows.
|
|
func calcColsMaximumNorm(mtxA [][]float64, c, r, n int) float64 {
|
|
var norm float64
|
|
for row := r; row < n; row++ {
|
|
if norm < math.Abs(mtxA[c][row]) {
|
|
norm = math.Abs(mtxA[c][row])
|
|
}
|
|
}
|
|
return norm
|
|
}
|
|
|
|
// calcFastMult returns multiply n x m matrix A with m x l matrix B to n x l matrix R.
|
|
func calcFastMult(mtxA, mtxB, mtxR [][]float64, n, m, l int) {
|
|
var sum float64
|
|
for row := 0; row < n; row++ {
|
|
for col := 0; col < l; col++ {
|
|
sum = 0.0
|
|
for k := 0; k < m; k++ {
|
|
sum += mtxA[k][row] * mtxB[col][k]
|
|
}
|
|
mtxR[col][row] = sum
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcRowsEuclideanNorm is a special version for use within QR
|
|
// decomposition. Euclidean norm of column index c starting in row index r;
|
|
// matrix a has count n rows.
|
|
func calcRowsEuclideanNorm(mtxA [][]float64, c, r, n int) float64 {
|
|
var norm float64
|
|
for row := r; row < n; row++ {
|
|
norm += mtxA[c][row] * mtxA[c][row]
|
|
}
|
|
return math.Sqrt(norm)
|
|
}
|
|
|
|
// calcRowsSumProduct is a special version for use within QR decomposition.
|
|
// <A(a);B(b)> starting in row index r;
|
|
// a and b are indices of columns, matrices A and B have count n rows.
|
|
func calcRowsSumProduct(mtxA [][]float64, a int, mtxB [][]float64, b, r, n int) float64 {
|
|
var result float64
|
|
for row := r; row < n; row++ {
|
|
result += mtxA[a][row] * mtxB[b][row]
|
|
}
|
|
return result
|
|
}
|
|
|
|
// calcSolveWithUpperRightTriangle solve for X in R*X=S using back substitution.
|
|
func calcSolveWithUpperRightTriangle(mtxA [][]float64, vecR []float64, mtxS [][]float64, k int, bIsTransposed bool) {
|
|
var row int
|
|
for rowp1 := k; rowp1 > 0; rowp1-- {
|
|
row = rowp1 - 1
|
|
sum := getDouble(mtxS, row)
|
|
for col := rowp1; col < k; col++ {
|
|
if bIsTransposed {
|
|
sum -= mtxA[row][col] * getDouble(mtxS, col)
|
|
} else {
|
|
sum -= mtxA[col][row] * getDouble(mtxS, col)
|
|
}
|
|
}
|
|
putDouble(mtxS, row, sum/vecR[row])
|
|
}
|
|
}
|
|
|
|
// calcRowQRDecomposition calculates a QR decomposition with Householder
|
|
// reflection.
|
|
func calcRowQRDecomposition(mtxA [][]float64, vecR []float64, k, n int) bool {
|
|
for col := 0; col < k; col++ {
|
|
scale := calcColsMaximumNorm(mtxA, col, col, n)
|
|
if scale == 0 {
|
|
return false
|
|
}
|
|
for row := col; row < n; row++ {
|
|
mtxA[col][row] = mtxA[col][row] / scale
|
|
}
|
|
euclid := calcRowsEuclideanNorm(mtxA, col, col, n)
|
|
factor := 1.0 / euclid / (euclid + math.Abs(mtxA[col][col]))
|
|
signum := calcSign(mtxA[col][col])
|
|
mtxA[col][col] = mtxA[col][col] + signum*euclid
|
|
vecR[col] = -signum * scale * euclid
|
|
// apply Householder transformation to A
|
|
for c := col + 1; c < k; c++ {
|
|
sum := calcRowsSumProduct(mtxA, col, mtxA, c, col, n)
|
|
for row := col; row < n; row++ {
|
|
mtxA[c][row] = mtxA[c][row] - sum*factor*mtxA[col][row]
|
|
}
|
|
}
|
|
}
|
|
return true
|
|
}
|
|
|
|
// calcApplyColsHouseholderTransformation transposed matrices A and Y.
|
|
func calcApplyColsHouseholderTransformation(mtxA [][]float64, r int, mtxY [][]float64, n int) {
|
|
denominator := calcColsSumProduct(mtxA, r, mtxA, r, r, n)
|
|
numerator := calcColsSumProduct(mtxA, r, mtxY, 0, r, n)
|
|
factor := 2 * (numerator / denominator)
|
|
for col := r; col < n; col++ {
|
|
putDouble(mtxY, col, getDouble(mtxY, col)-factor*mtxA[col][r])
|
|
}
|
|
}
|
|
|
|
// calcRowMeans calculates means of the rows of matrix.
|
|
func calcRowMeans(mtxX, mtxRes [][]float64, c, r int) {
|
|
for k := 0; k < r; k++ {
|
|
var fSum float64
|
|
for i := 0; i < c; i++ {
|
|
fSum += mtxX[i][k]
|
|
}
|
|
mtxRes[k][0] = fSum / float64(c)
|
|
}
|
|
}
|
|
|
|
// calcRowsDelta calculates subtract of the rows of matrix.
|
|
func calcRowsDelta(mtx, rowMeans [][]float64, c, r int) {
|
|
for k := 0; k < r; k++ {
|
|
for i := 0; i < c; i++ {
|
|
mtx[i][k] = approxSub(mtx[i][k], rowMeans[k][0])
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcColumnMaximumNorm returns maximum norm of row index R starting in col
|
|
// index C; matrix A has count N columns.
|
|
func calcColumnMaximumNorm(mtxA [][]float64, r, c, n int) float64 {
|
|
var norm float64
|
|
for col := c; col < n; col++ {
|
|
if norm < math.Abs(mtxA[col][r]) {
|
|
norm = math.Abs(mtxA[col][r])
|
|
}
|
|
}
|
|
return norm
|
|
}
|
|
|
|
// calcColsEuclideanNorm returns euclidean norm of row index R starting in
|
|
// column index C; matrix A has count N columns.
|
|
func calcColsEuclideanNorm(mtxA [][]float64, r, c, n int) float64 {
|
|
var norm float64
|
|
for col := c; col < n; col++ {
|
|
norm += (mtxA[col][r]) * (mtxA[col][r])
|
|
}
|
|
return math.Sqrt(norm)
|
|
}
|
|
|
|
// calcColsSumProduct returns sum product for given matrix.
|
|
func calcColsSumProduct(mtxA [][]float64, a int, mtxB [][]float64, b, c, n int) float64 {
|
|
var result float64
|
|
for col := c; col < n; col++ {
|
|
result += mtxA[col][a] * mtxB[col][b]
|
|
}
|
|
return result
|
|
}
|
|
|
|
// calcColQRDecomposition same with transposed matrix A, N is count of
|
|
// columns, k count of rows.
|
|
func calcColQRDecomposition(mtxA [][]float64, vecR []float64, k, n int) bool {
|
|
var sum float64
|
|
for row := 0; row < k; row++ {
|
|
// calculate vector u of the householder transformation
|
|
scale := calcColumnMaximumNorm(mtxA, row, row, n)
|
|
if scale == 0 {
|
|
return false
|
|
}
|
|
for col := row; col < n; col++ {
|
|
mtxA[col][row] = mtxA[col][row] / scale
|
|
}
|
|
euclid := calcColsEuclideanNorm(mtxA, row, row, n)
|
|
factor := 1 / euclid / (euclid + math.Abs(mtxA[row][row]))
|
|
signum := calcSign(mtxA[row][row])
|
|
mtxA[row][row] = mtxA[row][row] + signum*euclid
|
|
vecR[row] = -signum * scale * euclid
|
|
// apply Householder transformation to A
|
|
for r := row + 1; r < k; r++ {
|
|
sum = calcColsSumProduct(mtxA, row, mtxA, r, row, n)
|
|
for col := row; col < n; col++ {
|
|
mtxA[col][r] = mtxA[col][r] - sum*factor*mtxA[col][row]
|
|
}
|
|
}
|
|
}
|
|
return true
|
|
}
|
|
|
|
// calcApplyRowsHouseholderTransformation applies a Householder transformation to a
|
|
// column vector Y with is given as Nx1 Matrix. The vector u, from which the
|
|
// Householder transformation is built, is the column part in matrix A, with
|
|
// column index c, starting with row index c. A is the result of the QR
|
|
// decomposition as obtained from calcRowQRDecomposition.
|
|
func calcApplyRowsHouseholderTransformation(mtxA [][]float64, c int, mtxY [][]float64, n int) {
|
|
denominator := calcRowsSumProduct(mtxA, c, mtxA, c, c, n)
|
|
numerator := calcRowsSumProduct(mtxA, c, mtxY, 0, c, n)
|
|
factor := 2 * (numerator / denominator)
|
|
for row := c; row < n; row++ {
|
|
putDouble(mtxY, row, getDouble(mtxY, row)-factor*mtxA[c][row])
|
|
}
|
|
}
|
|
|
|
// calcTrendGrowthSimpleRegression calculate simple regression for the calcTrendGrowth.
|
|
func calcTrendGrowthSimpleRegression(bConstant, bGrowth bool, mtxY, mtxX, newX, mtxRes [][]float64, meanY float64, N int) {
|
|
var meanX float64
|
|
if bConstant {
|
|
meanX = calcMeanOverAll(mtxX, N)
|
|
for i := 0; i < len(mtxX); i++ {
|
|
for j := 0; j < len(mtxX[i]); j++ {
|
|
mtxX[i][j] = approxSub(mtxX[i][j], meanX)
|
|
}
|
|
}
|
|
}
|
|
sumXY := calcSumProduct(mtxX, mtxY, N)
|
|
sumX2 := calcSumProduct(mtxX, mtxX, N)
|
|
slope := sumXY / sumX2
|
|
var help float64
|
|
var intercept float64
|
|
if bConstant {
|
|
intercept = meanY - slope*meanX
|
|
for i := 0; i < len(mtxRes); i++ {
|
|
for j := 0; j < len(mtxRes[i]); j++ {
|
|
help = newX[i][j]*slope + intercept
|
|
if bGrowth {
|
|
mtxRes[i][j] = math.Exp(help)
|
|
} else {
|
|
mtxRes[i][j] = help
|
|
}
|
|
}
|
|
}
|
|
} else {
|
|
for i := 0; i < len(mtxRes); i++ {
|
|
for j := 0; j < len(mtxRes[i]); j++ {
|
|
help = newX[i][j] * slope
|
|
if bGrowth {
|
|
mtxRes[i][j] = math.Exp(help)
|
|
} else {
|
|
mtxRes[i][j] = help
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcTrendGrowthMultipleRegressionPart1 calculate multiple regression for the
|
|
// calcTrendGrowth.
|
|
func calcTrendGrowthMultipleRegressionPart1(bConstant, bGrowth bool, mtxY, mtxX, newX, mtxRes [][]float64, meanY float64, RXN, K, N int) {
|
|
vecR := make([]float64, N) // for QR decomposition
|
|
means := getNewMatrix(K, 1) // mean of each column
|
|
slopes := getNewMatrix(1, K) // from b1 to bK
|
|
if len(means) == 0 || len(slopes) == 0 {
|
|
return
|
|
}
|
|
if bConstant {
|
|
calcColumnMeans(mtxX, means, K, N)
|
|
calcColumnsDelta(mtxX, means, K, N)
|
|
}
|
|
if !calcRowQRDecomposition(mtxX, vecR, K, N) {
|
|
return
|
|
}
|
|
// Later on we will divide by elements of vecR, so make sure that they aren't zero.
|
|
bIsSingular := false
|
|
for row := 0; row < K && !bIsSingular; row++ {
|
|
bIsSingular = bIsSingular || vecR[row] == 0
|
|
}
|
|
if bIsSingular {
|
|
return
|
|
}
|
|
for col := 0; col < K; col++ {
|
|
calcApplyRowsHouseholderTransformation(mtxX, col, mtxY, N)
|
|
}
|
|
for col := 0; col < K; col++ {
|
|
putDouble(slopes, col, getDouble(mtxY, col))
|
|
}
|
|
calcSolveWithUpperRightTriangle(mtxX, vecR, slopes, K, false)
|
|
// Fill result matrix
|
|
calcFastMult(newX, slopes, mtxRes, RXN, K, 1)
|
|
if bConstant {
|
|
intercept := meanY - calcSumProduct(means, slopes, K)
|
|
for row := 0; row < RXN; row++ {
|
|
mtxRes[0][row] = mtxRes[0][row] + intercept
|
|
}
|
|
}
|
|
if bGrowth {
|
|
for i := 0; i < RXN; i++ {
|
|
putDouble(mtxRes, i, math.Exp(getDouble(mtxRes, i)))
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcTrendGrowthMultipleRegressionPart2 calculate multiple regression for the
|
|
// calcTrendGrowth.
|
|
func calcTrendGrowthMultipleRegressionPart2(bConstant, bGrowth bool, mtxY, mtxX, newX, mtxRes [][]float64, meanY float64, nCXN, K, N int) {
|
|
vecR := make([]float64, N) // for QR decomposition
|
|
means := getNewMatrix(K, 1) // mean of each row
|
|
slopes := getNewMatrix(K, 1) // row from b1 to bK
|
|
if len(means) == 0 || len(slopes) == 0 {
|
|
return
|
|
}
|
|
if bConstant {
|
|
calcRowMeans(mtxX, means, N, K)
|
|
calcRowsDelta(mtxX, means, N, K)
|
|
}
|
|
if !calcColQRDecomposition(mtxX, vecR, K, N) {
|
|
return
|
|
}
|
|
// later on we will divide by elements of vecR, so make sure that they aren't zero
|
|
bIsSingular := false
|
|
for row := 0; row < K && !bIsSingular; row++ {
|
|
bIsSingular = bIsSingular || vecR[row] == 0
|
|
}
|
|
if bIsSingular {
|
|
return
|
|
}
|
|
for row := 0; row < K; row++ {
|
|
calcApplyColsHouseholderTransformation(mtxX, row, mtxY, N)
|
|
}
|
|
for col := 0; col < K; col++ {
|
|
putDouble(slopes, col, getDouble(mtxY, col))
|
|
}
|
|
calcSolveWithUpperRightTriangle(mtxX, vecR, slopes, K, true)
|
|
// fill result matrix
|
|
calcFastMult(slopes, newX, mtxRes, 1, K, nCXN)
|
|
if bConstant {
|
|
fIntercept := meanY - calcSumProduct(means, slopes, K)
|
|
for col := 0; col < nCXN; col++ {
|
|
mtxRes[col][0] = mtxRes[col][0] + fIntercept
|
|
}
|
|
}
|
|
if bGrowth {
|
|
for i := 0; i < nCXN; i++ {
|
|
putDouble(mtxRes, i, math.Exp(getDouble(mtxRes, i)))
|
|
}
|
|
}
|
|
}
|
|
|
|
// calcTrendGrowthRegression is a part of implementation of the calcTrendGrowth.
|
|
func calcTrendGrowthRegression(bConstant, bGrowth bool, trendType, nCXN, nRXN, K, N int, mtxY, mtxX, newX, mtxRes [][]float64) {
|
|
if len(mtxRes) == 0 {
|
|
return
|
|
}
|
|
var meanY float64
|
|
if bConstant {
|
|
copyX, copyY := matrixClone(mtxX), matrixClone(mtxY)
|
|
mtxX, mtxY = copyX, copyY
|
|
meanY = calcMeanOverAll(mtxY, N)
|
|
for i := 0; i < len(mtxY); i++ {
|
|
for j := 0; j < len(mtxY[i]); j++ {
|
|
mtxY[i][j] = approxSub(mtxY[i][j], meanY)
|
|
}
|
|
}
|
|
}
|
|
switch trendType {
|
|
case 1:
|
|
calcTrendGrowthSimpleRegression(bConstant, bGrowth, mtxY, mtxX, newX, mtxRes, meanY, N)
|
|
case 2:
|
|
calcTrendGrowthMultipleRegressionPart1(bConstant, bGrowth, mtxY, mtxX, newX, mtxRes, meanY, nRXN, K, N)
|
|
default:
|
|
calcTrendGrowthMultipleRegressionPart2(bConstant, bGrowth, mtxY, mtxX, newX, mtxRes, meanY, nCXN, K, N)
|
|
}
|
|
}
|
|
|
|
// calcTrendGrowth returns values along a predicted exponential trend.
|
|
func calcTrendGrowth(mtxY, mtxX, newX [][]float64, bConstant, bGrowth bool) ([][]float64, formulaArg) {
|
|
getMatrixParams, errArg := prepareTrendGrowth(bGrowth, mtxX, mtxY)
|
|
if errArg.Type != ArgEmpty {
|
|
return nil, errArg
|
|
}
|
|
trendType := getMatrixParams.trendType
|
|
nCX := getMatrixParams.nCX
|
|
nRX := getMatrixParams.nRX
|
|
K := getMatrixParams.M
|
|
N := getMatrixParams.N
|
|
mtxX = getMatrixParams.mtxX
|
|
mtxY = getMatrixParams.mtxY
|
|
// checking if data samples are enough
|
|
if (bConstant && (N < K+1)) || (!bConstant && (N < K)) || (N < 1) || (K < 1) {
|
|
return nil, errArg
|
|
}
|
|
// set the default newX if necessary
|
|
nCXN, nRXN := nCX, nRX
|
|
if len(newX) == 0 {
|
|
newX = matrixClone(mtxX) // mtxX will be changed to X-meanX
|
|
} else {
|
|
nRXN, nCXN = len(newX[0]), len(newX)
|
|
if (trendType == 2 && K != nCXN) || (trendType == 3 && K != nRXN) {
|
|
return nil, errArg
|
|
}
|
|
}
|
|
var mtxRes [][]float64
|
|
switch trendType {
|
|
case 1:
|
|
mtxRes = getNewMatrix(nCXN, nRXN)
|
|
case 2:
|
|
mtxRes = getNewMatrix(1, nRXN)
|
|
default:
|
|
mtxRes = getNewMatrix(nCXN, 1)
|
|
}
|
|
calcTrendGrowthRegression(bConstant, bGrowth, trendType, nCXN, nRXN, K, N, mtxY, mtxX, newX, mtxRes)
|
|
return mtxRes, errArg
|
|
}
|
|
|
|
// trendGrowth is an implementation of the formula functions GROWTH and TREND.
|
|
func (fn *formulaFuncs) trendGrowth(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
if argsList.Len() > 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 4 arguments", name))
|
|
}
|
|
var knowY, knowX, newX [][]float64
|
|
var errArg formulaArg
|
|
constArg := newBoolFormulaArg(true)
|
|
knowY, errArg = newNumberMatrix(argsList.Front().Value.(formulaArg), false)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
if argsList.Len() > 1 {
|
|
knowX, errArg = newNumberMatrix(argsList.Front().Next().Value.(formulaArg), false)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
}
|
|
if argsList.Len() > 2 {
|
|
newX, errArg = newNumberMatrix(argsList.Front().Next().Next().Value.(formulaArg), false)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
}
|
|
if argsList.Len() > 3 {
|
|
if constArg = argsList.Back().Value.(formulaArg).ToBool(); constArg.Type != ArgNumber {
|
|
return constArg
|
|
}
|
|
}
|
|
var mtxNewX [][]float64
|
|
for i := 0; i < len(newX); i++ {
|
|
for j := 0; j < len(newX[i]); j++ {
|
|
for x := len(mtxNewX); x <= j; x++ {
|
|
mtxNewX = append(mtxNewX, []float64{})
|
|
}
|
|
for k := len(mtxNewX[j]); k <= i; k++ {
|
|
mtxNewX[j] = append(mtxNewX[j], 0)
|
|
}
|
|
mtxNewX[j][i] = newX[i][j]
|
|
}
|
|
}
|
|
mtx, errArg := calcTrendGrowth(knowY, knowX, mtxNewX, constArg.Number == 1, name == "GROWTH")
|
|
if errArg.Type != ArgEmpty {
|
|
return errArg
|
|
}
|
|
return newMatrixFormulaArg(newFormulaArgMatrix(mtx))
|
|
}
|
|
|
|
// GROWTH function calculates the exponential growth curve through a given set
|
|
// of y-values and (optionally), one or more sets of x-values. The function
|
|
// then extends the curve to calculate additional y-values for a further
|
|
// supplied set of new x-values. The syntax of the function is:
|
|
//
|
|
// GROWTH(known_y's,[known_x's],[new_x's],[const])
|
|
func (fn *formulaFuncs) GROWTH(argsList *list.List) formulaArg {
|
|
return fn.trendGrowth("GROWTH", argsList)
|
|
}
|
|
|
|
// HARMEAN function calculates the harmonic mean of a supplied set of values.
|
|
// The syntax of the function is:
|
|
//
|
|
// HARMEAN(number1,[number2],...)
|
|
func (fn *formulaFuncs) HARMEAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HARMEAN requires at least 1 argument")
|
|
}
|
|
if min := fn.MIN(argsList); min.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
number, val, cnt := 0.0, 0.0, 0.0
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
continue
|
|
}
|
|
number = num.Number
|
|
case ArgNumber:
|
|
number = arg.Number
|
|
}
|
|
if number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
val += 1 / number
|
|
cnt++
|
|
}
|
|
return newNumberFormulaArg(1 / (val / cnt))
|
|
}
|
|
|
|
// checkHYPGEOMDISTArgs checking arguments for the formula function HYPGEOMDIST
|
|
// and HYPGEOM.DIST.
|
|
func checkHYPGEOMDISTArgs(sampleS, numberSample, populationS, numberPop formulaArg) bool {
|
|
return sampleS.Number < 0 ||
|
|
sampleS.Number > math.Min(numberSample.Number, populationS.Number) ||
|
|
sampleS.Number < math.Max(0, numberSample.Number-numberPop.Number+populationS.Number) ||
|
|
numberSample.Number <= 0 ||
|
|
numberSample.Number > numberPop.Number ||
|
|
populationS.Number <= 0 ||
|
|
populationS.Number > numberPop.Number ||
|
|
numberPop.Number <= 0
|
|
}
|
|
|
|
// prepareHYPGEOMDISTArgs prepare arguments for the formula function
|
|
// HYPGEOMDIST and HYPGEOM.DIST.
|
|
func (fn *formulaFuncs) prepareHYPGEOMDISTArgs(name string, argsList *list.List) formulaArg {
|
|
if name == "HYPGEOMDIST" && argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HYPGEOMDIST requires 4 numeric arguments")
|
|
}
|
|
if name == "HYPGEOM.DIST" && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HYPGEOM.DIST requires 5 arguments")
|
|
}
|
|
var sampleS, numberSample, populationS, numberPop, cumulative formulaArg
|
|
if sampleS = argsList.Front().Value.(formulaArg).ToNumber(); sampleS.Type != ArgNumber {
|
|
return sampleS
|
|
}
|
|
if numberSample = argsList.Front().Next().Value.(formulaArg).ToNumber(); numberSample.Type != ArgNumber {
|
|
return numberSample
|
|
}
|
|
if populationS = argsList.Front().Next().Next().Value.(formulaArg).ToNumber(); populationS.Type != ArgNumber {
|
|
return populationS
|
|
}
|
|
if numberPop = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); numberPop.Type != ArgNumber {
|
|
return numberPop
|
|
}
|
|
if checkHYPGEOMDISTArgs(sampleS, numberSample, populationS, numberPop) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if name == "HYPGEOM.DIST" {
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type != ArgNumber {
|
|
return cumulative
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{sampleS, numberSample, populationS, numberPop, cumulative})
|
|
}
|
|
|
|
// HYPGEOMdotDIST function returns the value of the hypergeometric distribution
|
|
// for a specified number of successes from a population sample. The function
|
|
// can calculate the cumulative distribution or the probability density
|
|
// function. The syntax of the function is:
|
|
//
|
|
// HYPGEOM.DIST(sample_s,number_sample,population_s,number_pop,cumulative)
|
|
func (fn *formulaFuncs) HYPGEOMdotDIST(argsList *list.List) formulaArg {
|
|
args := fn.prepareHYPGEOMDISTArgs("HYPGEOM.DIST", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
sampleS, numberSample, populationS, numberPop, cumulative := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4]
|
|
if cumulative.Number == 1 {
|
|
var res float64
|
|
for i := 0; i <= int(sampleS.Number); i++ {
|
|
res += binomCoeff(populationS.Number, float64(i)) *
|
|
binomCoeff(numberPop.Number-populationS.Number, numberSample.Number-float64(i)) /
|
|
binomCoeff(numberPop.Number, numberSample.Number)
|
|
}
|
|
return newNumberFormulaArg(res)
|
|
}
|
|
return newNumberFormulaArg(binomCoeff(populationS.Number, sampleS.Number) *
|
|
binomCoeff(numberPop.Number-populationS.Number, numberSample.Number-sampleS.Number) /
|
|
binomCoeff(numberPop.Number, numberSample.Number))
|
|
}
|
|
|
|
// HYPGEOMDIST function returns the value of the hypergeometric distribution
|
|
// for a given number of successes from a sample of a population. The syntax
|
|
// of the function is:
|
|
//
|
|
// HYPGEOMDIST(sample_s,number_sample,population_s,number_pop)
|
|
func (fn *formulaFuncs) HYPGEOMDIST(argsList *list.List) formulaArg {
|
|
args := fn.prepareHYPGEOMDISTArgs("HYPGEOMDIST", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
sampleS, numberSample, populationS, numberPop := args.List[0], args.List[1], args.List[2], args.List[3]
|
|
return newNumberFormulaArg(binomCoeff(populationS.Number, sampleS.Number) *
|
|
binomCoeff(numberPop.Number-populationS.Number, numberSample.Number-sampleS.Number) /
|
|
binomCoeff(numberPop.Number, numberSample.Number))
|
|
}
|
|
|
|
// KURT function calculates the kurtosis of a supplied set of values. The
|
|
// syntax of the function is:
|
|
//
|
|
// KURT(number1,[number2],...)
|
|
func (fn *formulaFuncs) KURT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "KURT requires at least 1 argument")
|
|
}
|
|
mean, stdev := fn.AVERAGE(argsList), fn.STDEV(argsList)
|
|
if stdev.Number > 0 {
|
|
count, summer := 0.0, 0.0
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString, ArgNumber:
|
|
num := token.ToNumber()
|
|
if num.Type == ArgError {
|
|
continue
|
|
}
|
|
summer += math.Pow((num.Number-mean.Number)/stdev.Number, 4)
|
|
count++
|
|
case ArgList, ArgMatrix:
|
|
for _, row := range token.ToList() {
|
|
if row.Type == ArgNumber || row.Type == ArgString {
|
|
num := row.ToNumber()
|
|
if num.Type == ArgError {
|
|
continue
|
|
}
|
|
summer += math.Pow((num.Number-mean.Number)/stdev.Number, 4)
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if count > 3 {
|
|
return newNumberFormulaArg(summer*(count*(count+1)/((count-1)*(count-2)*(count-3))) - (3 * math.Pow(count-1, 2) / ((count - 2) * (count - 3))))
|
|
}
|
|
}
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
|
|
// EXPONdotDIST function returns the value of the exponential distribution for
|
|
// a give value of x. The user can specify whether the probability density
|
|
// function or the cumulative distribution function is used. The syntax of the
|
|
// Expondist function is:
|
|
//
|
|
// EXPON.DIST(x,lambda,cumulative)
|
|
func (fn *formulaFuncs) EXPONdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EXPON.DIST requires 3 arguments")
|
|
}
|
|
return fn.EXPONDIST(argsList)
|
|
}
|
|
|
|
// EXPONDIST function returns the value of the exponential distribution for a
|
|
// give value of x. The user can specify whether the probability density
|
|
// function or the cumulative distribution function is used. The syntax of the
|
|
// Expondist function is:
|
|
//
|
|
// EXPONDIST(x,lambda,cumulative)
|
|
func (fn *formulaFuncs) EXPONDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EXPONDIST requires 3 arguments")
|
|
}
|
|
var x, lambda, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if lambda = argsList.Front().Next().Value.(formulaArg).ToNumber(); lambda.Type != ArgNumber {
|
|
return lambda
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if x.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if lambda.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(1 - math.Exp(-lambda.Number*x.Number))
|
|
}
|
|
return newNumberFormulaArg(lambda.Number * math.Exp(-lambda.Number*x.Number))
|
|
}
|
|
|
|
// FdotDIST function calculates the Probability Density Function or the
|
|
// Cumulative Distribution Function for the F Distribution. This function is
|
|
// frequently used to measure the degree of diversity between two data
|
|
// sets. The syntax of the function is:
|
|
//
|
|
// F.DIST(x,deg_freedom1,deg_freedom2,cumulative)
|
|
func (fn *formulaFuncs) FdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "F.DIST requires 4 arguments")
|
|
}
|
|
var x, deg1, deg2, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if deg1 = argsList.Front().Next().Value.(formulaArg).ToNumber(); deg1.Type != ArgNumber {
|
|
return deg1
|
|
}
|
|
if deg2 = argsList.Front().Next().Next().Value.(formulaArg).ToNumber(); deg2.Type != ArgNumber {
|
|
return deg2
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if x.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
maxDeg := math.Pow10(10)
|
|
if deg1.Number < 1 || deg1.Number >= maxDeg {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if deg2.Number < 1 || deg2.Number >= maxDeg {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(1 - getBetaDist(deg2.Number/(deg2.Number+deg1.Number*x.Number), deg2.Number/2, deg1.Number/2))
|
|
}
|
|
return newNumberFormulaArg(math.Gamma((deg2.Number+deg1.Number)/2) / (math.Gamma(deg1.Number/2) * math.Gamma(deg2.Number/2)) * math.Pow(deg1.Number/deg2.Number, deg1.Number/2) * (math.Pow(x.Number, (deg1.Number-2)/2) / math.Pow(1+(deg1.Number/deg2.Number)*x.Number, (deg1.Number+deg2.Number)/2)))
|
|
}
|
|
|
|
// FDIST function calculates the (right-tailed) F Probability Distribution,
|
|
// which measures the degree of diversity between two data sets. The syntax
|
|
// of the function is:
|
|
//
|
|
// FDIST(x,deg_freedom1,deg_freedom2)
|
|
func (fn *formulaFuncs) FDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FDIST requires 3 arguments")
|
|
}
|
|
var x, deg1, deg2 formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if deg1 = argsList.Front().Next().Value.(formulaArg).ToNumber(); deg1.Type != ArgNumber {
|
|
return deg1
|
|
}
|
|
if deg2 = argsList.Back().Value.(formulaArg).ToNumber(); deg2.Type != ArgNumber {
|
|
return deg2
|
|
}
|
|
if x.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
maxDeg := math.Pow10(10)
|
|
if deg1.Number < 1 || deg1.Number >= maxDeg {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if deg2.Number < 1 || deg2.Number >= maxDeg {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg(deg1.Number * x.Number / (deg1.Number*x.Number + deg2.Number)))
|
|
args.PushBack(newNumberFormulaArg(0.5 * deg1.Number))
|
|
args.PushBack(newNumberFormulaArg(0.5 * deg2.Number))
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
return newNumberFormulaArg(1 - fn.BETADIST(args).Number)
|
|
}
|
|
|
|
// FdotDISTdotRT function calculates the (right-tailed) F Probability
|
|
// Distribution, which measures the degree of diversity between two data sets.
|
|
// The syntax of the function is:
|
|
//
|
|
// F.DIST.RT(x,deg_freedom1,deg_freedom2)
|
|
func (fn *formulaFuncs) FdotDISTdotRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "F.DIST.RT requires 3 arguments")
|
|
}
|
|
return fn.FDIST(argsList)
|
|
}
|
|
|
|
// prepareFinvArgs checking and prepare arguments for the formula function
|
|
// F.INV, F.INV.RT and FINV.
|
|
func (fn *formulaFuncs) prepareFinvArgs(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 3 arguments", name))
|
|
}
|
|
var probability, d1, d2 formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if d1 = argsList.Front().Next().Value.(formulaArg).ToNumber(); d1.Type != ArgNumber {
|
|
return d1
|
|
}
|
|
if d2 = argsList.Back().Value.(formulaArg).ToNumber(); d2.Type != ArgNumber {
|
|
return d2
|
|
}
|
|
if probability.Number <= 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if d1.Number < 1 || d1.Number >= math.Pow10(10) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if d2.Number < 1 || d2.Number >= math.Pow10(10) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newListFormulaArg([]formulaArg{probability, d1, d2})
|
|
}
|
|
|
|
// FdotINV function calculates the inverse of the Cumulative F Distribution
|
|
// for a supplied probability. The syntax of the F.Inv function is:
|
|
//
|
|
// F.INV(probability,deg_freedom1,deg_freedom2)
|
|
func (fn *formulaFuncs) FdotINV(argsList *list.List) formulaArg {
|
|
args := fn.prepareFinvArgs("F.INV", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
probability, d1, d2 := args.List[0], args.List[1], args.List[2]
|
|
return newNumberFormulaArg((1/calcBetainv(1-probability.Number, d2.Number/2, d1.Number/2, 0, 1) - 1) * (d2.Number / d1.Number))
|
|
}
|
|
|
|
// FdotINVdotRT function calculates the inverse of the (right-tailed) F
|
|
// Probability Distribution for a supplied probability. The syntax of the
|
|
// function is:
|
|
//
|
|
// F.INV.RT(probability,deg_freedom1,deg_freedom2)
|
|
func (fn *formulaFuncs) FdotINVdotRT(argsList *list.List) formulaArg {
|
|
args := fn.prepareFinvArgs("F.INV.RT", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
probability, d1, d2 := args.List[0], args.List[1], args.List[2]
|
|
return newNumberFormulaArg((1/calcBetainv(1-(1-probability.Number), d2.Number/2, d1.Number/2, 0, 1) - 1) * (d2.Number / d1.Number))
|
|
}
|
|
|
|
// FINV function calculates the inverse of the (right-tailed) F Probability
|
|
// Distribution for a supplied probability. The syntax of the function is:
|
|
//
|
|
// FINV(probability,deg_freedom1,deg_freedom2)
|
|
func (fn *formulaFuncs) FINV(argsList *list.List) formulaArg {
|
|
args := fn.prepareFinvArgs("FINV", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
probability, d1, d2 := args.List[0], args.List[1], args.List[2]
|
|
return newNumberFormulaArg((1/calcBetainv(1-(1-probability.Number), d2.Number/2, d1.Number/2, 0, 1) - 1) * (d2.Number / d1.Number))
|
|
}
|
|
|
|
// FdotTEST function returns the F-Test for two supplied arrays. I.e. the
|
|
// function returns the two-tailed probability that the variances in the two
|
|
// supplied arrays are not significantly different. The syntax of the Ftest
|
|
// function is:
|
|
//
|
|
// F.TEST(array1,array2)
|
|
func (fn *formulaFuncs) FdotTEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "F.TEST requires 2 arguments")
|
|
}
|
|
array1 := argsList.Front().Value.(formulaArg)
|
|
array2 := argsList.Back().Value.(formulaArg)
|
|
left, right := array1.ToList(), array2.ToList()
|
|
collectMatrix := func(args []formulaArg) (n, accu float64) {
|
|
var p, sum float64
|
|
for _, arg := range args {
|
|
if num := arg.ToNumber(); num.Type == ArgNumber {
|
|
x := num.Number - p
|
|
y := x / (n + 1)
|
|
p += y
|
|
accu += n * x * y
|
|
n++
|
|
sum += num.Number
|
|
}
|
|
}
|
|
return
|
|
}
|
|
nums, accu := collectMatrix(left)
|
|
f3 := nums - 1
|
|
if nums == 1 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
f1 := accu / (nums - 1)
|
|
if f1 == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
nums, accu = collectMatrix(right)
|
|
f4 := nums - 1
|
|
if nums == 1 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
f2 := accu / (nums - 1)
|
|
if f2 == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg(f1 / f2))
|
|
args.PushBack(newNumberFormulaArg(f3))
|
|
args.PushBack(newNumberFormulaArg(f4))
|
|
probability := (1 - fn.FDIST(args).Number) * 2
|
|
if probability > 1 {
|
|
probability = 2 - probability
|
|
}
|
|
return newNumberFormulaArg(probability)
|
|
}
|
|
|
|
// FTEST function returns the F-Test for two supplied arrays. I.e. the function
|
|
// returns the two-tailed probability that the variances in the two supplied
|
|
// arrays are not significantly different. The syntax of the Ftest function
|
|
// is:
|
|
//
|
|
// FTEST(array1,array2)
|
|
func (fn *formulaFuncs) FTEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FTEST requires 2 arguments")
|
|
}
|
|
return fn.FdotTEST(argsList)
|
|
}
|
|
|
|
// LOGINV function calculates the inverse of the Cumulative Log-Normal
|
|
// Distribution Function of x, for a supplied probability. The syntax of the
|
|
// function is:
|
|
//
|
|
// LOGINV(probability,mean,standard_dev)
|
|
func (fn *formulaFuncs) LOGINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOGINV requires 3 arguments")
|
|
}
|
|
var probability, mean, stdDev formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if stdDev = argsList.Back().Value.(formulaArg).ToNumber(); stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if probability.Number <= 0 || probability.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if stdDev.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(probability)
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
norminv := fn.NORMINV(args)
|
|
return newNumberFormulaArg(math.Exp(mean.Number + stdDev.Number*norminv.Number))
|
|
}
|
|
|
|
// LOGNORMdotINV function calculates the inverse of the Cumulative Log-Normal
|
|
// Distribution Function of x, for a supplied probability. The syntax of the
|
|
// function is:
|
|
//
|
|
// LOGNORM.INV(probability,mean,standard_dev)
|
|
func (fn *formulaFuncs) LOGNORMdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOGNORM.INV requires 3 arguments")
|
|
}
|
|
return fn.LOGINV(argsList)
|
|
}
|
|
|
|
// LOGNORMdotDIST function calculates the Log-Normal Probability Density
|
|
// Function or the Cumulative Log-Normal Distribution Function for a supplied
|
|
// value of x. The syntax of the function is:
|
|
//
|
|
// LOGNORM.DIST(x,mean,standard_dev,cumulative)
|
|
func (fn *formulaFuncs) LOGNORMdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOGNORM.DIST requires 4 arguments")
|
|
}
|
|
var x, mean, stdDev, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if stdDev = argsList.Back().Prev().Value.(formulaArg).ToNumber(); stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if x.Number <= 0 || stdDev.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg((math.Log(x.Number) - mean.Number) / stdDev.Number))
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
args.PushBack(cumulative)
|
|
return fn.NORMDIST(args)
|
|
}
|
|
return newNumberFormulaArg((1 / (math.Sqrt(2*math.Pi) * stdDev.Number * x.Number)) *
|
|
math.Exp(0-(math.Pow(math.Log(x.Number)-mean.Number, 2)/(2*math.Pow(stdDev.Number, 2)))))
|
|
}
|
|
|
|
// LOGNORMDIST function calculates the Cumulative Log-Normal Distribution
|
|
// Function at a supplied value of x. The syntax of the function is:
|
|
//
|
|
// LOGNORMDIST(x,mean,standard_dev)
|
|
func (fn *formulaFuncs) LOGNORMDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOGNORMDIST requires 3 arguments")
|
|
}
|
|
var x, mean, stdDev formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if stdDev = argsList.Back().Value.(formulaArg).ToNumber(); stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if x.Number <= 0 || stdDev.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg((math.Log(x.Number) - mean.Number) / stdDev.Number))
|
|
return fn.NORMSDIST(args)
|
|
}
|
|
|
|
// MODE function returns the statistical mode (the most frequently occurring
|
|
// value) of a list of supplied numbers. If there are 2 or more most
|
|
// frequently occurring values in the supplied data, the function returns the
|
|
// lowest of these values The syntax of the function is:
|
|
//
|
|
// MODE(number1,[number2],...)
|
|
func (fn *formulaFuncs) MODE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MODE requires at least 1 argument")
|
|
}
|
|
var values []float64
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
cells := arg.Value.(formulaArg)
|
|
if cells.Type != ArgMatrix && cells.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
for _, cell := range cells.ToList() {
|
|
if cell.Type == ArgNumber {
|
|
values = append(values, cell.Number)
|
|
}
|
|
}
|
|
}
|
|
sort.Float64s(values)
|
|
cnt := len(values)
|
|
var count, modeCnt int
|
|
var mode float64
|
|
for i := 0; i < cnt; i++ {
|
|
count = 0
|
|
for j := 0; j < cnt; j++ {
|
|
if j != i && values[j] == values[i] {
|
|
count++
|
|
}
|
|
}
|
|
if count > modeCnt {
|
|
modeCnt = count
|
|
mode = values[i]
|
|
}
|
|
}
|
|
if modeCnt == 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(mode)
|
|
}
|
|
|
|
// MODEdotMULT function returns a vertical array of the statistical modes
|
|
// (the most frequently occurring values) within a list of supplied numbers.
|
|
// The syntax of the function is:
|
|
//
|
|
// MODE.MULT(number1,[number2],...)
|
|
func (fn *formulaFuncs) MODEdotMULT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MODE.MULT requires at least 1 argument")
|
|
}
|
|
var values []float64
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
cells := arg.Value.(formulaArg)
|
|
if cells.Type != ArgMatrix && cells.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
for _, cell := range cells.ToList() {
|
|
if cell.Type == ArgNumber {
|
|
values = append(values, cell.Number)
|
|
}
|
|
}
|
|
}
|
|
sort.Float64s(values)
|
|
cnt := len(values)
|
|
var count, modeCnt int
|
|
var mtx [][]formulaArg
|
|
for i := 0; i < cnt; i++ {
|
|
count = 0
|
|
for j := i + 1; j < cnt; j++ {
|
|
if values[i] == values[j] {
|
|
count++
|
|
}
|
|
}
|
|
if count > modeCnt {
|
|
modeCnt = count
|
|
mtx = [][]formulaArg{}
|
|
mtx = append(mtx, []formulaArg{newNumberFormulaArg(values[i])})
|
|
} else if count == modeCnt {
|
|
mtx = append(mtx, []formulaArg{newNumberFormulaArg(values[i])})
|
|
}
|
|
}
|
|
if modeCnt == 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newMatrixFormulaArg(mtx)
|
|
}
|
|
|
|
// MODEdotSNGL function returns the statistical mode (the most frequently
|
|
// occurring value) within a list of supplied numbers. If there are 2 or more
|
|
// most frequently occurring values in the supplied data, the function returns
|
|
// the lowest of these values. The syntax of the function is:
|
|
//
|
|
// MODE.SNGL(number1,[number2],...)
|
|
func (fn *formulaFuncs) MODEdotSNGL(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MODE.SNGL requires at least 1 argument")
|
|
}
|
|
return fn.MODE(argsList)
|
|
}
|
|
|
|
// NEGBINOMdotDIST function calculates the probability mass function or the
|
|
// cumulative distribution function for the Negative Binomial Distribution.
|
|
// This gives the probability that there will be a given number of failures
|
|
// before a required number of successes is achieved. The syntax of the
|
|
// function is:
|
|
//
|
|
// NEGBINOM.DIST(number_f,number_s,probability_s,cumulative)
|
|
func (fn *formulaFuncs) NEGBINOMdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NEGBINOM.DIST requires 4 arguments")
|
|
}
|
|
var f, s, probability, cumulative formulaArg
|
|
if f = argsList.Front().Value.(formulaArg).ToNumber(); f.Type != ArgNumber {
|
|
return f
|
|
}
|
|
if s = argsList.Front().Next().Value.(formulaArg).ToNumber(); s.Type != ArgNumber {
|
|
return s
|
|
}
|
|
if probability = argsList.Front().Next().Next().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type != ArgNumber {
|
|
return cumulative
|
|
}
|
|
if f.Number < 0 || s.Number < 1 || probability.Number < 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(1 - getBetaDist(1-probability.Number, f.Number+1, s.Number))
|
|
}
|
|
return newNumberFormulaArg(binomCoeff(f.Number+s.Number-1, s.Number-1) * math.Pow(probability.Number, s.Number) * math.Pow(1-probability.Number, f.Number))
|
|
}
|
|
|
|
// NEGBINOMDIST function calculates the Negative Binomial Distribution for a
|
|
// given set of parameters. This gives the probability that there will be a
|
|
// specified number of failures before a required number of successes is
|
|
// achieved. The syntax of the function is:
|
|
//
|
|
// NEGBINOMDIST(number_f,number_s,probability_s)
|
|
func (fn *formulaFuncs) NEGBINOMDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NEGBINOMDIST requires 3 arguments")
|
|
}
|
|
var f, s, probability formulaArg
|
|
if f = argsList.Front().Value.(formulaArg).ToNumber(); f.Type != ArgNumber {
|
|
return f
|
|
}
|
|
if s = argsList.Front().Next().Value.(formulaArg).ToNumber(); s.Type != ArgNumber {
|
|
return s
|
|
}
|
|
if probability = argsList.Back().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if f.Number < 0 || s.Number < 1 || probability.Number < 0 || probability.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(binomCoeff(f.Number+s.Number-1, s.Number-1) * math.Pow(probability.Number, s.Number) * math.Pow(1-probability.Number, f.Number))
|
|
}
|
|
|
|
// NORMdotDIST function calculates the Normal Probability Density Function or
|
|
// the Cumulative Normal Distribution. Function for a supplied set of
|
|
// parameters. The syntax of the function is:
|
|
//
|
|
// NORM.DIST(x,mean,standard_dev,cumulative)
|
|
func (fn *formulaFuncs) NORMdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORM.DIST requires 4 arguments")
|
|
}
|
|
return fn.NORMDIST(argsList)
|
|
}
|
|
|
|
// NORMDIST function calculates the Normal Probability Density Function or the
|
|
// Cumulative Normal Distribution. Function for a supplied set of parameters.
|
|
// The syntax of the function is:
|
|
//
|
|
// NORMDIST(x,mean,standard_dev,cumulative)
|
|
func (fn *formulaFuncs) NORMDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORMDIST requires 4 arguments")
|
|
}
|
|
var x, mean, stdDev, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if stdDev = argsList.Back().Prev().Value.(formulaArg).ToNumber(); stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type == ArgError {
|
|
return cumulative
|
|
}
|
|
if stdDev.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if cumulative.Number == 1 {
|
|
return newNumberFormulaArg(0.5 * (1 + math.Erf((x.Number-mean.Number)/(stdDev.Number*math.Sqrt(2)))))
|
|
}
|
|
return newNumberFormulaArg((1 / (math.Sqrt(2*math.Pi) * stdDev.Number)) * math.Exp(0-(math.Pow(x.Number-mean.Number, 2)/(2*(stdDev.Number*stdDev.Number)))))
|
|
}
|
|
|
|
// NORMdotINV function calculates the inverse of the Cumulative Normal
|
|
// Distribution Function for a supplied value of x, and a supplied
|
|
// distribution mean & standard deviation. The syntax of the function is:
|
|
//
|
|
// NORM.INV(probability,mean,standard_dev)
|
|
func (fn *formulaFuncs) NORMdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORM.INV requires 3 arguments")
|
|
}
|
|
return fn.NORMINV(argsList)
|
|
}
|
|
|
|
// NORMINV function calculates the inverse of the Cumulative Normal
|
|
// Distribution Function for a supplied value of x, and a supplied
|
|
// distribution mean & standard deviation. The syntax of the function is:
|
|
//
|
|
// NORMINV(probability,mean,standard_dev)
|
|
func (fn *formulaFuncs) NORMINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORMINV requires 3 arguments")
|
|
}
|
|
var prob, mean, stdDev formulaArg
|
|
if prob = argsList.Front().Value.(formulaArg).ToNumber(); prob.Type != ArgNumber {
|
|
return prob
|
|
}
|
|
if mean = argsList.Front().Next().Value.(formulaArg).ToNumber(); mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
if stdDev = argsList.Back().Value.(formulaArg).ToNumber(); stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if prob.Number < 0 || prob.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if stdDev.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
inv, err := norminv(prob.Number)
|
|
if err != nil {
|
|
return newErrorFormulaArg(err.Error(), err.Error())
|
|
}
|
|
return newNumberFormulaArg(inv*stdDev.Number + mean.Number)
|
|
}
|
|
|
|
// NORMdotSdotDIST function calculates the Standard Normal Cumulative
|
|
// Distribution Function for a supplied value. The syntax of the function
|
|
// is:
|
|
//
|
|
// NORM.S.DIST(z)
|
|
func (fn *formulaFuncs) NORMdotSdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORM.S.DIST requires 2 numeric arguments")
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 0})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1})
|
|
args.PushBack(argsList.Back().Value.(formulaArg))
|
|
return fn.NORMDIST(args)
|
|
}
|
|
|
|
// NORMSDIST function calculates the Standard Normal Cumulative Distribution
|
|
// Function for a supplied value. The syntax of the function is:
|
|
//
|
|
// NORMSDIST(z)
|
|
func (fn *formulaFuncs) NORMSDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORMSDIST requires 1 numeric argument")
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 0})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1, Boolean: true})
|
|
return fn.NORMDIST(args)
|
|
}
|
|
|
|
// NORMSINV function calculates the inverse of the Standard Normal Cumulative
|
|
// Distribution Function for a supplied probability value. The syntax of the
|
|
// function is:
|
|
//
|
|
// NORMSINV(probability)
|
|
func (fn *formulaFuncs) NORMSINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORMSINV requires 1 numeric argument")
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 0})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1})
|
|
return fn.NORMINV(args)
|
|
}
|
|
|
|
// NORMdotSdotINV function calculates the inverse of the Standard Normal
|
|
// Cumulative Distribution Function for a supplied probability value. The
|
|
// syntax of the function is:
|
|
//
|
|
// NORM.S.INV(probability)
|
|
func (fn *formulaFuncs) NORMdotSdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NORM.S.INV requires 1 numeric argument")
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 0})
|
|
args.PushBack(formulaArg{Type: ArgNumber, Number: 1})
|
|
return fn.NORMINV(args)
|
|
}
|
|
|
|
// norminv returns the inverse of the normal cumulative distribution for the
|
|
// specified value.
|
|
func norminv(p float64) (float64, error) {
|
|
a := map[int]float64{
|
|
1: -3.969683028665376e+01, 2: 2.209460984245205e+02, 3: -2.759285104469687e+02,
|
|
4: 1.383577518672690e+02, 5: -3.066479806614716e+01, 6: 2.506628277459239e+00,
|
|
}
|
|
b := map[int]float64{
|
|
1: -5.447609879822406e+01, 2: 1.615858368580409e+02, 3: -1.556989798598866e+02,
|
|
4: 6.680131188771972e+01, 5: -1.328068155288572e+01,
|
|
}
|
|
c := map[int]float64{
|
|
1: -7.784894002430293e-03, 2: -3.223964580411365e-01, 3: -2.400758277161838e+00,
|
|
4: -2.549732539343734e+00, 5: 4.374664141464968e+00, 6: 2.938163982698783e+00,
|
|
}
|
|
d := map[int]float64{
|
|
1: 7.784695709041462e-03, 2: 3.224671290700398e-01, 3: 2.445134137142996e+00,
|
|
4: 3.754408661907416e+00,
|
|
}
|
|
pLow := 0.02425 // Use lower region approx. below this
|
|
pHigh := 1 - pLow // Use upper region approx. above this
|
|
if 0 < p && p < pLow {
|
|
// Rational approximation for lower region.
|
|
q := math.Sqrt(-2 * math.Log(p))
|
|
return (((((c[1]*q+c[2])*q+c[3])*q+c[4])*q+c[5])*q + c[6]) /
|
|
((((d[1]*q+d[2])*q+d[3])*q+d[4])*q + 1), nil
|
|
} else if pLow <= p && p <= pHigh {
|
|
// Rational approximation for central region.
|
|
q := p - 0.5
|
|
r := q * q
|
|
f1 := ((((a[1]*r+a[2])*r+a[3])*r+a[4])*r + a[5]) * r
|
|
f2 := (b[1]*r + b[2]) * r
|
|
f3 := ((math.Nextafter(f2, f2)+b[3])*r + b[4]) * r
|
|
f4 := (math.Nextafter(f3, f3) + b[5]) * r
|
|
return (math.Nextafter(f1, f1) + a[6]) * q /
|
|
(math.Nextafter(f4, f4) + 1), nil
|
|
} else if pHigh < p && p < 1 {
|
|
// Rational approximation for upper region.
|
|
q := math.Sqrt(-2 * math.Log(1-p))
|
|
return -(((((c[1]*q+c[2])*q+c[3])*q+c[4])*q+c[5])*q + c[6]) /
|
|
((((d[1]*q+d[2])*q+d[3])*q+d[4])*q + 1), nil
|
|
}
|
|
return 0, errors.New(formulaErrorNUM)
|
|
}
|
|
|
|
// kth is an implementation of the formula functions LARGE and SMALL.
|
|
func (fn *formulaFuncs) kth(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 arguments", name))
|
|
}
|
|
array := argsList.Front().Value.(formulaArg).ToList()
|
|
argK := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if argK.Type != ArgNumber {
|
|
return argK
|
|
}
|
|
k := int(argK.Number)
|
|
if k < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "k should be > 0")
|
|
}
|
|
var data []float64
|
|
for _, arg := range array {
|
|
if arg.Type == ArgNumber {
|
|
data = append(data, arg.Number)
|
|
}
|
|
}
|
|
if len(data) < k {
|
|
return newErrorFormulaArg(formulaErrorNUM, "k should be <= length of array")
|
|
}
|
|
sort.Float64s(data)
|
|
if name == "LARGE" {
|
|
return newNumberFormulaArg(data[len(data)-k])
|
|
}
|
|
return newNumberFormulaArg(data[k-1])
|
|
}
|
|
|
|
// LARGE function returns the k'th largest value from an array of numeric
|
|
// values. The syntax of the function is:
|
|
//
|
|
// LARGE(array,k)
|
|
func (fn *formulaFuncs) LARGE(argsList *list.List) formulaArg {
|
|
return fn.kth("LARGE", argsList)
|
|
}
|
|
|
|
// MAX function returns the largest value from a supplied set of numeric
|
|
// values. The syntax of the function is:
|
|
//
|
|
// MAX(number1,[number2],...)
|
|
func (fn *formulaFuncs) MAX(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MAX requires at least 1 argument")
|
|
}
|
|
return fn.max(false, argsList)
|
|
}
|
|
|
|
// MAXA function returns the largest value from a supplied set of numeric
|
|
// values, while counting text and the logical value FALSE as the value 0 and
|
|
// counting the logical value TRUE as the value 1. The syntax of the function
|
|
// is:
|
|
//
|
|
// MAXA(number1,[number2],...)
|
|
func (fn *formulaFuncs) MAXA(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MAXA requires at least 1 argument")
|
|
}
|
|
return fn.max(true, argsList)
|
|
}
|
|
|
|
// MAXIFS function returns the maximum value from a subset of values that are
|
|
// specified according to one or more criteria. The syntax of the function
|
|
// is:
|
|
//
|
|
// MAXIFS(max_range,criteria_range1,criteria1,[criteria_range2,criteria2],...)
|
|
func (fn *formulaFuncs) MAXIFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MAXIFS requires at least 3 arguments")
|
|
}
|
|
if argsList.Len()%2 != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var args []formulaArg
|
|
max, maxRange := -math.MaxFloat64, argsList.Front().Value.(formulaArg).Matrix
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
for _, ref := range formulaIfsMatch(args) {
|
|
if num := maxRange[ref.Row][ref.Col].ToNumber(); num.Type == ArgNumber && max < num.Number {
|
|
max = num.Number
|
|
}
|
|
}
|
|
if max == -math.MaxFloat64 {
|
|
max = 0
|
|
}
|
|
return newNumberFormulaArg(max)
|
|
}
|
|
|
|
// calcListMatrixMax is part of the implementation max.
|
|
func calcListMatrixMax(maxa bool, max float64, arg formulaArg) float64 {
|
|
for _, cell := range arg.ToList() {
|
|
if cell.Type == ArgNumber && cell.Number > max {
|
|
if maxa && cell.Boolean || !cell.Boolean {
|
|
max = cell.Number
|
|
}
|
|
}
|
|
}
|
|
return max
|
|
}
|
|
|
|
// max is an implementation of the formula functions MAX and MAXA.
|
|
func (fn *formulaFuncs) max(maxa bool, argsList *list.List) formulaArg {
|
|
max := -math.MaxFloat64
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
if !maxa && (arg.Value() == "TRUE" || arg.Value() == "FALSE") {
|
|
continue
|
|
} else {
|
|
num := arg.ToBool()
|
|
if num.Type == ArgNumber && num.Number > max {
|
|
max = num.Number
|
|
continue
|
|
}
|
|
}
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgError && num.Number > max {
|
|
max = num.Number
|
|
}
|
|
case ArgNumber:
|
|
if arg.Number > max {
|
|
max = arg.Number
|
|
}
|
|
case ArgList, ArgMatrix:
|
|
max = calcListMatrixMax(maxa, max, arg)
|
|
case ArgError:
|
|
return arg
|
|
}
|
|
}
|
|
if max == -math.MaxFloat64 {
|
|
max = 0
|
|
}
|
|
return newNumberFormulaArg(max)
|
|
}
|
|
|
|
// MEDIAN function returns the statistical median (the middle value) of a list
|
|
// of supplied numbers. The syntax of the function is:
|
|
//
|
|
// MEDIAN(number1,[number2],...)
|
|
func (fn *formulaFuncs) MEDIAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MEDIAN requires at least 1 argument")
|
|
}
|
|
var values []float64
|
|
var median float64
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
value := arg.ToNumber()
|
|
if value.Type != ArgNumber {
|
|
return value
|
|
}
|
|
values = append(values, value.Number)
|
|
case ArgNumber:
|
|
values = append(values, arg.Number)
|
|
case ArgMatrix:
|
|
for _, row := range arg.Matrix {
|
|
for _, cell := range row {
|
|
if cell.Type == ArgNumber {
|
|
values = append(values, cell.Number)
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if len(values) == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
sort.Float64s(values)
|
|
if len(values)%2 == 0 {
|
|
median = (values[len(values)/2-1] + values[len(values)/2]) / 2
|
|
} else {
|
|
median = values[len(values)/2]
|
|
}
|
|
return newNumberFormulaArg(median)
|
|
}
|
|
|
|
// MIN function returns the smallest value from a supplied set of numeric
|
|
// values. The syntax of the function is:
|
|
//
|
|
// MIN(number1,[number2],...)
|
|
func (fn *formulaFuncs) MIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MIN requires at least 1 argument")
|
|
}
|
|
return fn.min(false, argsList)
|
|
}
|
|
|
|
// MINA function returns the smallest value from a supplied set of numeric
|
|
// values, while counting text and the logical value FALSE as the value 0 and
|
|
// counting the logical value TRUE as the value 1. The syntax of the function
|
|
// is:
|
|
//
|
|
// MINA(number1,[number2],...)
|
|
func (fn *formulaFuncs) MINA(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MINA requires at least 1 argument")
|
|
}
|
|
return fn.min(true, argsList)
|
|
}
|
|
|
|
// MINIFS function returns the minimum value from a subset of values that are
|
|
// specified according to one or more criteria. The syntax of the function
|
|
// is:
|
|
//
|
|
// MINIFS(min_range,criteria_range1,criteria1,[criteria_range2,criteria2],...)
|
|
func (fn *formulaFuncs) MINIFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MINIFS requires at least 3 arguments")
|
|
}
|
|
if argsList.Len()%2 != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var args []formulaArg
|
|
min, minRange := math.MaxFloat64, argsList.Front().Value.(formulaArg).Matrix
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
args = append(args, arg.Value.(formulaArg))
|
|
}
|
|
for _, ref := range formulaIfsMatch(args) {
|
|
if num := minRange[ref.Row][ref.Col].ToNumber(); num.Type == ArgNumber && min > num.Number {
|
|
min = num.Number
|
|
}
|
|
}
|
|
if min == math.MaxFloat64 {
|
|
min = 0
|
|
}
|
|
return newNumberFormulaArg(min)
|
|
}
|
|
|
|
// calcListMatrixMin is part of the implementation min.
|
|
func calcListMatrixMin(mina bool, min float64, arg formulaArg) float64 {
|
|
for _, cell := range arg.ToList() {
|
|
if cell.Type == ArgNumber && cell.Number < min {
|
|
if mina && cell.Boolean || !cell.Boolean {
|
|
min = cell.Number
|
|
}
|
|
}
|
|
}
|
|
return min
|
|
}
|
|
|
|
// min is an implementation of the formula functions MIN and MINA.
|
|
func (fn *formulaFuncs) min(mina bool, argsList *list.List) formulaArg {
|
|
min := math.MaxFloat64
|
|
for token := argsList.Front(); token != nil; token = token.Next() {
|
|
arg := token.Value.(formulaArg)
|
|
switch arg.Type {
|
|
case ArgString:
|
|
if !mina && (arg.Value() == "TRUE" || arg.Value() == "FALSE") {
|
|
continue
|
|
} else {
|
|
num := arg.ToBool()
|
|
if num.Type == ArgNumber && num.Number < min {
|
|
min = num.Number
|
|
continue
|
|
}
|
|
}
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgError && num.Number < min {
|
|
min = num.Number
|
|
}
|
|
case ArgNumber:
|
|
if arg.Number < min {
|
|
min = arg.Number
|
|
}
|
|
case ArgList, ArgMatrix:
|
|
min = calcListMatrixMin(mina, min, arg)
|
|
case ArgError:
|
|
return arg
|
|
}
|
|
}
|
|
if min == math.MaxFloat64 {
|
|
min = 0
|
|
}
|
|
return newNumberFormulaArg(min)
|
|
}
|
|
|
|
// pearsonProduct is an implementation of the formula functions PEARSON, RSQ
|
|
// and SLOPE.
|
|
func (fn *formulaFuncs) pearsonProduct(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 arguments", name))
|
|
}
|
|
var array1, array2 []formulaArg
|
|
if name == "SLOPE" {
|
|
array1 = argsList.Back().Value.(formulaArg).ToList()
|
|
array2 = argsList.Front().Value.(formulaArg).ToList()
|
|
} else {
|
|
array1 = argsList.Front().Value.(formulaArg).ToList()
|
|
array2 = argsList.Back().Value.(formulaArg).ToList()
|
|
}
|
|
if len(array1) != len(array2) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var sum, deltaX, deltaY, x, y, length float64
|
|
for i := 0; i < len(array1); i++ {
|
|
num1, num2 := array1[i], array2[i]
|
|
if !(num1.Type == ArgNumber && num2.Type == ArgNumber) {
|
|
continue
|
|
}
|
|
x += num1.Number
|
|
y += num2.Number
|
|
length++
|
|
}
|
|
x /= length
|
|
y /= length
|
|
for i := 0; i < len(array1); i++ {
|
|
num1, num2 := array1[i], array2[i]
|
|
if !(num1.Type == ArgNumber && num2.Type == ArgNumber) {
|
|
continue
|
|
}
|
|
sum += (num1.Number - x) * (num2.Number - y)
|
|
deltaX += (num1.Number - x) * (num1.Number - x)
|
|
deltaY += (num2.Number - y) * (num2.Number - y)
|
|
}
|
|
if deltaX == 0 || deltaY == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
if name == "RSQ" {
|
|
return newNumberFormulaArg(math.Pow(sum/math.Sqrt(deltaX*deltaY), 2))
|
|
}
|
|
if name == "PEARSON" {
|
|
return newNumberFormulaArg(sum / math.Sqrt(deltaX*deltaY))
|
|
}
|
|
return newNumberFormulaArg(sum / deltaX)
|
|
}
|
|
|
|
// PEARSON function calculates the Pearson Product-Moment Correlation
|
|
// Coefficient for two sets of values. The syntax of the function is:
|
|
//
|
|
// PEARSON(array1,array2)
|
|
func (fn *formulaFuncs) PEARSON(argsList *list.List) formulaArg {
|
|
return fn.pearsonProduct("PEARSON", argsList)
|
|
}
|
|
|
|
// PERCENTILEdotEXC function returns the k'th percentile (i.e. the value below
|
|
// which k% of the data values fall) for a supplied range of values and a
|
|
// supplied k (between 0 & 1 exclusive).The syntax of the function is:
|
|
//
|
|
// PERCENTILE.EXC(array,k)
|
|
func (fn *formulaFuncs) PERCENTILEdotEXC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PERCENTILE.EXC requires 2 arguments")
|
|
}
|
|
array := argsList.Front().Value.(formulaArg).ToList()
|
|
k := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if k.Type != ArgNumber {
|
|
return k
|
|
}
|
|
if k.Number <= 0 || k.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
var numbers []float64
|
|
for _, arg := range array {
|
|
if arg.Type == ArgError {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if arg.Type == ArgNumber {
|
|
numbers = append(numbers, arg.Number)
|
|
}
|
|
}
|
|
cnt := len(numbers)
|
|
sort.Float64s(numbers)
|
|
idx := k.Number * (float64(cnt) + 1)
|
|
base := math.Floor(idx)
|
|
next := base - 1
|
|
proportion := math.Nextafter(idx, idx) - base
|
|
return newNumberFormulaArg(numbers[int(next)] + ((numbers[int(base)] - numbers[int(next)]) * proportion))
|
|
}
|
|
|
|
// PERCENTILEdotINC function returns the k'th percentile (i.e. the value below
|
|
// which k% of the data values fall) for a supplied range of values and a
|
|
// supplied k. The syntax of the function is:
|
|
//
|
|
// PERCENTILE.INC(array,k)
|
|
func (fn *formulaFuncs) PERCENTILEdotINC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PERCENTILE.INC requires 2 arguments")
|
|
}
|
|
return fn.PERCENTILE(argsList)
|
|
}
|
|
|
|
// PERCENTILE function returns the k'th percentile (i.e. the value below which
|
|
// k% of the data values fall) for a supplied range of values and a supplied
|
|
// k. The syntax of the function is:
|
|
//
|
|
// PERCENTILE(array,k)
|
|
func (fn *formulaFuncs) PERCENTILE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PERCENTILE requires 2 arguments")
|
|
}
|
|
array := argsList.Front().Value.(formulaArg).ToList()
|
|
k := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if k.Type != ArgNumber {
|
|
return k
|
|
}
|
|
if k.Number < 0 || k.Number > 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var numbers []float64
|
|
for _, arg := range array {
|
|
if arg.Type == ArgError {
|
|
return arg
|
|
}
|
|
if arg.Type == ArgNumber {
|
|
numbers = append(numbers, arg.Number)
|
|
}
|
|
}
|
|
cnt := len(numbers)
|
|
sort.Float64s(numbers)
|
|
idx := k.Number * (float64(cnt) - 1)
|
|
base := math.Floor(idx)
|
|
if idx == base {
|
|
return newNumberFormulaArg(numbers[int(idx)])
|
|
}
|
|
next := base + 1
|
|
proportion := math.Nextafter(idx, idx) - base
|
|
return newNumberFormulaArg(numbers[int(base)] + ((numbers[int(next)] - numbers[int(base)]) * proportion))
|
|
}
|
|
|
|
// percentrank is an implementation of the formula functions PERCENTRANK and
|
|
// PERCENTRANK.INC.
|
|
func (fn *formulaFuncs) percentrank(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 && argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 or 3 arguments", name))
|
|
}
|
|
array := argsList.Front().Value.(formulaArg).ToList()
|
|
x := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
var numbers []float64
|
|
for _, arg := range array {
|
|
if arg.Type == ArgError {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if arg.Type == ArgNumber {
|
|
numbers = append(numbers, arg.Number)
|
|
}
|
|
}
|
|
cnt := len(numbers)
|
|
sort.Float64s(numbers)
|
|
if x.Number < numbers[0] || x.Number > numbers[cnt-1] {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
pos, significance := float64(inFloat64Slice(numbers, x.Number)), newNumberFormulaArg(3)
|
|
if argsList.Len() == 3 {
|
|
if significance = argsList.Back().Value.(formulaArg).ToNumber(); significance.Type != ArgNumber {
|
|
return significance
|
|
}
|
|
if significance.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, fmt.Sprintf("%s arguments significance should be > 1", name))
|
|
}
|
|
}
|
|
if pos == -1 {
|
|
pos = 0
|
|
cmp := numbers[0]
|
|
for cmp < x.Number {
|
|
pos++
|
|
cmp = numbers[int(pos)]
|
|
}
|
|
pos--
|
|
pos += (x.Number - numbers[int(pos)]) / (cmp - numbers[int(pos)])
|
|
}
|
|
pow := math.Pow(10, significance.Number)
|
|
digit := pow * pos / (float64(cnt) - 1)
|
|
if name == "PERCENTRANK.EXC" {
|
|
digit = pow * (pos + 1) / (float64(cnt) + 1)
|
|
}
|
|
return newNumberFormulaArg(math.Floor(digit) / pow)
|
|
}
|
|
|
|
// PERCENTRANKdotEXC function calculates the relative position, between 0 and
|
|
// 1 (exclusive), of a specified value within a supplied array. The syntax of
|
|
// the function is:
|
|
//
|
|
// PERCENTRANK.EXC(array,x,[significance])
|
|
func (fn *formulaFuncs) PERCENTRANKdotEXC(argsList *list.List) formulaArg {
|
|
return fn.percentrank("PERCENTRANK.EXC", argsList)
|
|
}
|
|
|
|
// PERCENTRANKdotINC function calculates the relative position, between 0 and
|
|
// 1 (inclusive), of a specified value within a supplied array.The syntax of
|
|
// the function is:
|
|
//
|
|
// PERCENTRANK.INC(array,x,[significance])
|
|
func (fn *formulaFuncs) PERCENTRANKdotINC(argsList *list.List) formulaArg {
|
|
return fn.percentrank("PERCENTRANK.INC", argsList)
|
|
}
|
|
|
|
// PERCENTRANK function calculates the relative position of a specified value,
|
|
// within a set of values, as a percentage. The syntax of the function is:
|
|
//
|
|
// PERCENTRANK(array,x,[significance])
|
|
func (fn *formulaFuncs) PERCENTRANK(argsList *list.List) formulaArg {
|
|
return fn.percentrank("PERCENTRANK", argsList)
|
|
}
|
|
|
|
// PERMUT function calculates the number of permutations of a specified number
|
|
// of objects from a set of objects. The syntax of the function is:
|
|
//
|
|
// PERMUT(number,number_chosen)
|
|
func (fn *formulaFuncs) PERMUT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PERMUT requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
chosen := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return number
|
|
}
|
|
if chosen.Type != ArgNumber {
|
|
return chosen
|
|
}
|
|
if number.Number < chosen.Number {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(math.Round(fact(number.Number) / fact(number.Number-chosen.Number)))
|
|
}
|
|
|
|
// PERMUTATIONA function calculates the number of permutations, with
|
|
// repetitions, of a specified number of objects from a set. The syntax of
|
|
// the function is:
|
|
//
|
|
// PERMUTATIONA(number,number_chosen)
|
|
func (fn *formulaFuncs) PERMUTATIONA(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PERMUTATIONA requires 2 numeric arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
chosen := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return number
|
|
}
|
|
if chosen.Type != ArgNumber {
|
|
return chosen
|
|
}
|
|
num, numChosen := math.Floor(number.Number), math.Floor(chosen.Number)
|
|
if num < 0 || numChosen < 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(math.Pow(num, numChosen))
|
|
}
|
|
|
|
// PHI function returns the value of the density function for a standard normal
|
|
// distribution for a supplied number. The syntax of the function is:
|
|
//
|
|
// PHI(x)
|
|
func (fn *formulaFuncs) PHI(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PHI requires 1 argument")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
return newNumberFormulaArg(0.39894228040143268 * math.Exp(-(x.Number*x.Number)/2))
|
|
}
|
|
|
|
// QUARTILE function returns a requested quartile of a supplied range of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// QUARTILE(array,quart)
|
|
func (fn *formulaFuncs) QUARTILE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "QUARTILE requires 2 arguments")
|
|
}
|
|
quart := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if quart.Type != ArgNumber {
|
|
return quart
|
|
}
|
|
if quart.Number < 0 || quart.Number > 4 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(newNumberFormulaArg(quart.Number / 4))
|
|
return fn.PERCENTILE(args)
|
|
}
|
|
|
|
// QUARTILEdotEXC function returns a requested quartile of a supplied range of
|
|
// values, based on a percentile range of 0 to 1 exclusive. The syntax of the
|
|
// function is:
|
|
//
|
|
// QUARTILE.EXC(array,quart)
|
|
func (fn *formulaFuncs) QUARTILEdotEXC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "QUARTILE.EXC requires 2 arguments")
|
|
}
|
|
quart := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if quart.Type != ArgNumber {
|
|
return quart
|
|
}
|
|
if quart.Number <= 0 || quart.Number >= 4 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(newNumberFormulaArg(quart.Number / 4))
|
|
return fn.PERCENTILEdotEXC(args)
|
|
}
|
|
|
|
// QUARTILEdotINC function returns a requested quartile of a supplied range of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// QUARTILE.INC(array,quart)
|
|
func (fn *formulaFuncs) QUARTILEdotINC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "QUARTILE.INC requires 2 arguments")
|
|
}
|
|
return fn.QUARTILE(argsList)
|
|
}
|
|
|
|
// rank is an implementation of the formula functions RANK and RANK.EQ.
|
|
func (fn *formulaFuncs) rank(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 2 arguments", name))
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at most 3 arguments", name))
|
|
}
|
|
num := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return num
|
|
}
|
|
var arr []float64
|
|
for _, arg := range argsList.Front().Next().Value.(formulaArg).ToList() {
|
|
if arg.Type == ArgNumber {
|
|
arr = append(arr, arg.Number)
|
|
}
|
|
}
|
|
sort.Float64s(arr)
|
|
order := newNumberFormulaArg(0)
|
|
if argsList.Len() == 3 {
|
|
if order = argsList.Back().Value.(formulaArg).ToNumber(); order.Type != ArgNumber {
|
|
return order
|
|
}
|
|
}
|
|
if order.Number == 0 {
|
|
sort.Sort(sort.Reverse(sort.Float64Slice(arr)))
|
|
}
|
|
if idx := inFloat64Slice(arr, num.Number); idx != -1 {
|
|
return newNumberFormulaArg(float64(idx + 1))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// RANKdotEQ function returns the statistical rank of a given value, within a
|
|
// supplied array of values. If there are duplicate values in the list, these
|
|
// are given the same rank. The syntax of the function is:
|
|
//
|
|
// RANK.EQ(number,ref,[order])
|
|
func (fn *formulaFuncs) RANKdotEQ(argsList *list.List) formulaArg {
|
|
return fn.rank("RANK.EQ", argsList)
|
|
}
|
|
|
|
// RANK function returns the statistical rank of a given value, within a
|
|
// supplied array of values. If there are duplicate values in the list, these
|
|
// are given the same rank. The syntax of the function is:
|
|
//
|
|
// RANK(number,ref,[order])
|
|
func (fn *formulaFuncs) RANK(argsList *list.List) formulaArg {
|
|
return fn.rank("RANK", argsList)
|
|
}
|
|
|
|
// RSQ function calculates the square of the Pearson Product-Moment Correlation
|
|
// Coefficient for two supplied sets of values. The syntax of the function
|
|
// is:
|
|
//
|
|
// RSQ(known_y's,known_x's)
|
|
func (fn *formulaFuncs) RSQ(argsList *list.List) formulaArg {
|
|
return fn.pearsonProduct("RSQ", argsList)
|
|
}
|
|
|
|
// skew is an implementation of the formula functions SKEW and SKEW.P.
|
|
func (fn *formulaFuncs) skew(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
mean := fn.AVERAGE(argsList)
|
|
var stdDev formulaArg
|
|
var count, summer float64
|
|
if name == "SKEW" {
|
|
stdDev = fn.STDEV(argsList)
|
|
} else {
|
|
stdDev = fn.STDEVP(argsList)
|
|
}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgNumber, ArgString:
|
|
num := token.ToNumber()
|
|
if num.Type == ArgError {
|
|
return num
|
|
}
|
|
summer += math.Pow((num.Number-mean.Number)/stdDev.Number, 3)
|
|
count++
|
|
case ArgList, ArgMatrix:
|
|
for _, cell := range token.ToList() {
|
|
if cell.Type != ArgNumber {
|
|
continue
|
|
}
|
|
summer += math.Pow((cell.Number-mean.Number)/stdDev.Number, 3)
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
if count > 2 {
|
|
if name == "SKEW" {
|
|
return newNumberFormulaArg(summer * (count / ((count - 1) * (count - 2))))
|
|
}
|
|
return newNumberFormulaArg(summer / count)
|
|
}
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
|
|
// SKEW function calculates the skewness of the distribution of a supplied set
|
|
// of values. The syntax of the function is:
|
|
//
|
|
// SKEW(number1,[number2],...)
|
|
func (fn *formulaFuncs) SKEW(argsList *list.List) formulaArg {
|
|
return fn.skew("SKEW", argsList)
|
|
}
|
|
|
|
// SKEWdotP function calculates the skewness of the distribution of a supplied
|
|
// set of values. The syntax of the function is:
|
|
//
|
|
// SKEW.P(number1,[number2],...)
|
|
func (fn *formulaFuncs) SKEWdotP(argsList *list.List) formulaArg {
|
|
return fn.skew("SKEW.P", argsList)
|
|
}
|
|
|
|
// SLOPE returns the slope of the linear regression line through data points in
|
|
// known_y's and known_x's. The slope is the vertical distance divided by the
|
|
// horizontal distance between any two points on the line, which is the rate
|
|
// of change along the regression line. The syntax of the function is:
|
|
//
|
|
// SLOPE(known_y's,known_x's)
|
|
func (fn *formulaFuncs) SLOPE(argsList *list.List) formulaArg {
|
|
return fn.pearsonProduct("SLOPE", argsList)
|
|
}
|
|
|
|
// SMALL function returns the k'th smallest value from an array of numeric
|
|
// values. The syntax of the function is:
|
|
//
|
|
// SMALL(array,k)
|
|
func (fn *formulaFuncs) SMALL(argsList *list.List) formulaArg {
|
|
return fn.kth("SMALL", argsList)
|
|
}
|
|
|
|
// STANDARDIZE function returns a normalized value of a distribution that is
|
|
// characterized by a supplied mean and standard deviation. The syntax of the
|
|
// function is:
|
|
//
|
|
// STANDARDIZE(x,mean,standard_dev)
|
|
func (fn *formulaFuncs) STANDARDIZE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "STANDARDIZE requires 3 arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
mean := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if mean.Type != ArgNumber {
|
|
return mean
|
|
}
|
|
stdDev := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if stdDev.Type != ArgNumber {
|
|
return stdDev
|
|
}
|
|
if stdDev.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg((x.Number - mean.Number) / stdDev.Number)
|
|
}
|
|
|
|
// stdevp is an implementation of the formula functions STDEVP, STDEV.P and
|
|
// STDEVPA.
|
|
func (fn *formulaFuncs) stdevp(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
fnName := "VARP"
|
|
if name == "STDEVPA" {
|
|
fnName = "VARPA"
|
|
}
|
|
varp := fn.vars(fnName, argsList)
|
|
if varp.Type != ArgNumber {
|
|
return varp
|
|
}
|
|
return newNumberFormulaArg(math.Sqrt(varp.Number))
|
|
}
|
|
|
|
// STDEVP function calculates the standard deviation of a supplied set of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// STDEVP(number1,[number2],...)
|
|
func (fn *formulaFuncs) STDEVP(argsList *list.List) formulaArg {
|
|
return fn.stdevp("STDEVP", argsList)
|
|
}
|
|
|
|
// STDEVdotP function calculates the standard deviation of a supplied set of
|
|
// values.
|
|
//
|
|
// STDEV.P( number1, [number2], ... )
|
|
func (fn *formulaFuncs) STDEVdotP(argsList *list.List) formulaArg {
|
|
return fn.stdevp("STDEV.P", argsList)
|
|
}
|
|
|
|
// STDEVPA function calculates the standard deviation of a supplied set of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// STDEVPA(number1,[number2],...)
|
|
func (fn *formulaFuncs) STDEVPA(argsList *list.List) formulaArg {
|
|
return fn.stdevp("STDEVPA", argsList)
|
|
}
|
|
|
|
// STEYX function calculates the standard error for the line of best fit,
|
|
// through a supplied set of x- and y- values. The syntax of the function is:
|
|
//
|
|
// STEYX(known_y's,known_x's)
|
|
func (fn *formulaFuncs) STEYX(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "STEYX requires 2 arguments")
|
|
}
|
|
array1 := argsList.Back().Value.(formulaArg).ToList()
|
|
array2 := argsList.Front().Value.(formulaArg).ToList()
|
|
if len(array1) != len(array2) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var count, sumX, sumY, squareX, squareY, sigmaXY float64
|
|
for i := 0; i < len(array1); i++ {
|
|
num1, num2 := array1[i], array2[i]
|
|
if !(num1.Type == ArgNumber && num2.Type == ArgNumber) {
|
|
continue
|
|
}
|
|
sumX += num1.Number
|
|
sumY += num2.Number
|
|
squareX += num1.Number * num1.Number
|
|
squareY += num2.Number * num2.Number
|
|
sigmaXY += num1.Number * num2.Number
|
|
count++
|
|
}
|
|
if count < 3 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
dx, dy := sumX/count, sumY/count
|
|
sigma1 := squareY - 2*dy*sumY + count*dy*dy
|
|
sigma2 := sigmaXY - dy*sumX - sumY*dx + count*dy*dx
|
|
sigma3 := squareX - 2*dx*sumX + count*dx*dx
|
|
return newNumberFormulaArg(math.Sqrt((sigma1 - (sigma2*sigma2)/sigma3) / (count - 2)))
|
|
}
|
|
|
|
// getTDist is an implementation for the beta distribution probability density
|
|
// function.
|
|
func getTDist(T, fDF, nType float64) float64 {
|
|
var res float64
|
|
switch nType {
|
|
case 1:
|
|
res = 0.5 * getBetaDist(fDF/(fDF+T*T), fDF/2, 0.5)
|
|
case 2:
|
|
res = getBetaDist(fDF/(fDF+T*T), fDF/2, 0.5)
|
|
case 3:
|
|
res = math.Pow(1+(T*T/fDF), -(fDF+1)/2) / (math.Sqrt(fDF) * getBeta(0.5, fDF/2.0))
|
|
case 4:
|
|
X := fDF / (T*T + fDF)
|
|
R := 0.5 * getBetaDist(X, 0.5*fDF, 0.5)
|
|
res = 1 - R
|
|
if T < 0 {
|
|
res = R
|
|
}
|
|
}
|
|
return res
|
|
}
|
|
|
|
// TdotDIST function calculates the one-tailed Student's T Distribution, which
|
|
// is a continuous probability distribution that is frequently used for
|
|
// testing hypotheses on small sample data sets. The syntax of the function
|
|
// is:
|
|
//
|
|
// T.DIST(x,degrees_freedom,cumulative)
|
|
func (fn *formulaFuncs) TdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.DIST requires 3 arguments")
|
|
}
|
|
var x, degrees, cumulative formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if degrees = argsList.Front().Next().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if cumulative = argsList.Back().Value.(formulaArg).ToBool(); cumulative.Type != ArgNumber {
|
|
return cumulative
|
|
}
|
|
if cumulative.Number == 1 && degrees.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if cumulative.Number == 0 {
|
|
if degrees.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if degrees.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(getTDist(x.Number, degrees.Number, 3))
|
|
}
|
|
return newNumberFormulaArg(getTDist(x.Number, degrees.Number, 4))
|
|
}
|
|
|
|
// TdotDISTdot2T function calculates the two-tailed Student's T Distribution,
|
|
// which is a continuous probability distribution that is frequently used for
|
|
// testing hypotheses on small sample data sets. The syntax of the function
|
|
// is:
|
|
//
|
|
// T.DIST.2T(x,degrees_freedom)
|
|
func (fn *formulaFuncs) TdotDISTdot2T(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.DIST.2T requires 2 arguments")
|
|
}
|
|
var x, degrees formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if degrees = argsList.Back().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if x.Number < 0 || degrees.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(getTDist(x.Number, degrees.Number, 2))
|
|
}
|
|
|
|
// TdotDISTdotRT function calculates the right-tailed Student's T Distribution,
|
|
// which is a continuous probability distribution that is frequently used for
|
|
// testing hypotheses on small sample data sets. The syntax of the function
|
|
// is:
|
|
//
|
|
// T.DIST.RT(x,degrees_freedom)
|
|
func (fn *formulaFuncs) TdotDISTdotRT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.DIST.RT requires 2 arguments")
|
|
}
|
|
var x, degrees formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if degrees = argsList.Back().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if degrees.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
v := getTDist(x.Number, degrees.Number, 1)
|
|
if x.Number < 0 {
|
|
v = 1 - v
|
|
}
|
|
return newNumberFormulaArg(v)
|
|
}
|
|
|
|
// TDIST function calculates the Student's T Distribution, which is a
|
|
// continuous probability distribution that is frequently used for testing
|
|
// hypotheses on small sample data sets. The syntax of the function is:
|
|
//
|
|
// TDIST(x,degrees_freedom,tails)
|
|
func (fn *formulaFuncs) TDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TDIST requires 3 arguments")
|
|
}
|
|
var x, degrees, tails formulaArg
|
|
if x = argsList.Front().Value.(formulaArg).ToNumber(); x.Type != ArgNumber {
|
|
return x
|
|
}
|
|
if degrees = argsList.Front().Next().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if tails = argsList.Back().Value.(formulaArg).ToNumber(); tails.Type != ArgNumber {
|
|
return tails
|
|
}
|
|
if x.Number < 0 || degrees.Number < 1 || (tails.Number != 1 && tails.Number != 2) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(getTDist(x.Number, degrees.Number, tails.Number))
|
|
}
|
|
|
|
// TdotINV function calculates the left-tailed inverse of the Student's T
|
|
// Distribution, which is a continuous probability distribution that is
|
|
// frequently used for testing hypotheses on small sample data sets. The
|
|
// syntax of the function is:
|
|
//
|
|
// T.INV(probability,degrees_freedom)
|
|
func (fn *formulaFuncs) TdotINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.INV requires 2 arguments")
|
|
}
|
|
var probability, degrees formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if degrees = argsList.Back().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if probability.Number <= 0 || probability.Number >= 1 || degrees.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if probability.Number < 0.5 {
|
|
return newNumberFormulaArg(-calcIterateInverse(calcInverseIterator{
|
|
name: "T.INV",
|
|
fp: 1 - probability.Number,
|
|
fDF: degrees.Number,
|
|
nT: 4,
|
|
}, degrees.Number/2, degrees.Number))
|
|
}
|
|
return newNumberFormulaArg(calcIterateInverse(calcInverseIterator{
|
|
name: "T.INV",
|
|
fp: probability.Number,
|
|
fDF: degrees.Number,
|
|
nT: 4,
|
|
}, degrees.Number/2, degrees.Number))
|
|
}
|
|
|
|
// TdotINVdot2T function calculates the inverse of the two-tailed Student's T
|
|
// Distribution, which is a continuous probability distribution that is
|
|
// frequently used for testing hypotheses on small sample data sets. The
|
|
// syntax of the function is:
|
|
//
|
|
// T.INV.2T(probability,degrees_freedom)
|
|
func (fn *formulaFuncs) TdotINVdot2T(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.INV.2T requires 2 arguments")
|
|
}
|
|
var probability, degrees formulaArg
|
|
if probability = argsList.Front().Value.(formulaArg).ToNumber(); probability.Type != ArgNumber {
|
|
return probability
|
|
}
|
|
if degrees = argsList.Back().Value.(formulaArg).ToNumber(); degrees.Type != ArgNumber {
|
|
return degrees
|
|
}
|
|
if probability.Number <= 0 || probability.Number > 1 || degrees.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(calcIterateInverse(calcInverseIterator{
|
|
name: "T.INV.2T",
|
|
fp: probability.Number,
|
|
fDF: degrees.Number,
|
|
nT: 2,
|
|
}, degrees.Number/2, degrees.Number))
|
|
}
|
|
|
|
// TINV function calculates the inverse of the two-tailed Student's T
|
|
// Distribution, which is a continuous probability distribution that is
|
|
// frequently used for testing hypotheses on small sample data sets. The
|
|
// syntax of the function is:
|
|
//
|
|
// TINV(probability,degrees_freedom)
|
|
func (fn *formulaFuncs) TINV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TINV requires 2 arguments")
|
|
}
|
|
return fn.TdotINVdot2T(argsList)
|
|
}
|
|
|
|
// TREND function calculates the linear trend line through a given set of
|
|
// y-values and (optionally), a given set of x-values. The function then
|
|
// extends the linear trendline to calculate additional y-values for a further
|
|
// supplied set of new x-values. The syntax of the function is:
|
|
//
|
|
// TREND(known_y's,[known_x's],[new_x's],[const])
|
|
func (fn *formulaFuncs) TREND(argsList *list.List) formulaArg {
|
|
return fn.trendGrowth("TREND", argsList)
|
|
}
|
|
|
|
// tTest calculates the probability associated with the Student's T Test.
|
|
func tTest(bTemplin bool, mtx1, mtx2 [][]formulaArg, c1, c2, r1, r2 int) (float64, float64, bool) {
|
|
var cnt1, cnt2, sum1, sumSqr1, sum2, sumSqr2 float64
|
|
var fVal formulaArg
|
|
for i := 0; i < c1; i++ {
|
|
for j := 0; j < r1; j++ {
|
|
if fVal = mtx1[i][j]; fVal.Type == ArgNumber {
|
|
sum1 += fVal.Number
|
|
sumSqr1 += fVal.Number * fVal.Number
|
|
cnt1++
|
|
}
|
|
}
|
|
}
|
|
for i := 0; i < c2; i++ {
|
|
for j := 0; j < r2; j++ {
|
|
if fVal = mtx2[i][j]; fVal.Type == ArgNumber {
|
|
sum2 += fVal.Number
|
|
sumSqr2 += fVal.Number * fVal.Number
|
|
cnt2++
|
|
}
|
|
}
|
|
}
|
|
if cnt1 < 2.0 || cnt2 < 2.0 {
|
|
return 0, 0, false
|
|
}
|
|
if bTemplin {
|
|
fS1 := (sumSqr1 - sum1*sum1/cnt1) / (cnt1 - 1) / cnt1
|
|
fS2 := (sumSqr2 - sum2*sum2/cnt2) / (cnt2 - 1) / cnt2
|
|
if fS1+fS2 == 0 {
|
|
return 0, 0, false
|
|
}
|
|
c := fS1 / (fS1 + fS2)
|
|
return math.Abs(sum1/cnt1-sum2/cnt2) / math.Sqrt(fS1+fS2), 1 / (c*c/(cnt1-1) + (1-c)*(1-c)/(cnt2-1)), true
|
|
}
|
|
fS1 := (sumSqr1 - sum1*sum1/cnt1) / (cnt1 - 1)
|
|
fS2 := (sumSqr2 - sum2*sum2/cnt2) / (cnt2 - 1)
|
|
return math.Abs(sum1/cnt1-sum2/cnt2) / math.Sqrt((cnt1-1)*fS1+(cnt2-1)*fS2) * math.Sqrt(cnt1*cnt2*(cnt1+cnt2-2)/(cnt1+cnt2)), cnt1 + cnt2 - 2, true
|
|
}
|
|
|
|
// tTest is an implementation of the formula function TTEST.
|
|
func (fn *formulaFuncs) tTest(mtx1, mtx2 [][]formulaArg, fTails, fTyp float64) formulaArg {
|
|
var fT, fF float64
|
|
c1, c2, r1, r2, ok := len(mtx1), len(mtx2), len(mtx1[0]), len(mtx2[0]), true
|
|
if fTyp == 1 {
|
|
if c1 != c2 || r1 != r2 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
var cnt, sum1, sum2, sumSqrD float64
|
|
var fVal1, fVal2 formulaArg
|
|
for i := 0; i < c1; i++ {
|
|
for j := 0; j < r1; j++ {
|
|
fVal1, fVal2 = mtx1[i][j], mtx2[i][j]
|
|
if fVal1.Type != ArgNumber || fVal2.Type != ArgNumber {
|
|
continue
|
|
}
|
|
sum1 += fVal1.Number
|
|
sum2 += fVal2.Number
|
|
sumSqrD += (fVal1.Number - fVal2.Number) * (fVal1.Number - fVal2.Number)
|
|
cnt++
|
|
}
|
|
}
|
|
if cnt < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
sumD := sum1 - sum2
|
|
divider := cnt*sumSqrD - sumD*sumD
|
|
if divider == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
fT = math.Abs(sumD) * math.Sqrt((cnt-1)/divider)
|
|
fF = cnt - 1
|
|
} else if fTyp == 2 {
|
|
fT, fF, ok = tTest(false, mtx1, mtx2, c1, c2, r1, r2)
|
|
} else {
|
|
fT, fF, ok = tTest(true, mtx1, mtx2, c1, c2, r1, r2)
|
|
}
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(getTDist(fT, fF, fTails))
|
|
}
|
|
|
|
// TTEST function calculates the probability associated with the Student's T
|
|
// Test, which is commonly used for identifying whether two data sets are
|
|
// likely to have come from the same two underlying populations with the same
|
|
// mean. The syntax of the function is:
|
|
//
|
|
// TTEST(array1,array2,tails,type)
|
|
func (fn *formulaFuncs) TTEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TTEST requires 4 arguments")
|
|
}
|
|
var array1, array2, tails, typeArg formulaArg
|
|
array1 = argsList.Front().Value.(formulaArg)
|
|
array2 = argsList.Front().Next().Value.(formulaArg)
|
|
if tails = argsList.Front().Next().Next().Value.(formulaArg); tails.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if typeArg = argsList.Back().Value.(formulaArg); typeArg.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if len(array1.Matrix) == 0 || len(array2.Matrix) == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if tails.Number != 1 && tails.Number != 2 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if typeArg.Number != 1 && typeArg.Number != 2 && typeArg.Number != 3 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return fn.tTest(array1.Matrix, array2.Matrix, tails.Number, typeArg.Number)
|
|
}
|
|
|
|
// TdotTEST function calculates the probability associated with the Student's T
|
|
// Test, which is commonly used for identifying whether two data sets are
|
|
// likely to have come from the same two underlying populations with the same
|
|
// mean. The syntax of the function is:
|
|
//
|
|
// T.TEST(array1,array2,tails,type)
|
|
func (fn *formulaFuncs) TdotTEST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T.TEST requires 4 arguments")
|
|
}
|
|
return fn.TTEST(argsList)
|
|
}
|
|
|
|
// TRIMMEAN function calculates the trimmed mean (or truncated mean) of a
|
|
// supplied set of values. The syntax of the function is:
|
|
//
|
|
// TRIMMEAN(array,percent)
|
|
func (fn *formulaFuncs) TRIMMEAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TRIMMEAN requires 2 arguments")
|
|
}
|
|
percent := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if percent.Type != ArgNumber {
|
|
return percent
|
|
}
|
|
if percent.Number < 0 || percent.Number >= 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
var arr []float64
|
|
arrArg := argsList.Front().Value.(formulaArg).ToList()
|
|
for _, cell := range arrArg {
|
|
if cell.Type != ArgNumber {
|
|
continue
|
|
}
|
|
arr = append(arr, cell.Number)
|
|
}
|
|
discard := math.Floor(float64(len(arr)) * percent.Number / 2)
|
|
sort.Float64s(arr)
|
|
for i := 0; i < int(discard); i++ {
|
|
if len(arr) > 0 {
|
|
arr = arr[1:]
|
|
}
|
|
if len(arr) > 0 {
|
|
arr = arr[:len(arr)-1]
|
|
}
|
|
}
|
|
|
|
args := list.New().Init()
|
|
for _, ele := range arr {
|
|
args.PushBack(newNumberFormulaArg(ele))
|
|
}
|
|
return fn.AVERAGE(args)
|
|
}
|
|
|
|
// vars is an implementation of the formula functions VAR, VARA, VARP, VAR.P
|
|
// VAR.S and VARPA.
|
|
func (fn *formulaFuncs) vars(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
summerA, summerB, count := 0.0, 0.0, 0.0
|
|
minimum := 0.0
|
|
if name == "VAR" || name == "VAR.S" || name == "VARA" {
|
|
minimum = 1.0
|
|
}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
for _, token := range arg.Value.(formulaArg).ToList() {
|
|
if token.Value() == "" {
|
|
continue
|
|
}
|
|
num := token.ToNumber()
|
|
if token.Value() != "TRUE" && num.Type == ArgNumber {
|
|
summerA += num.Number * num.Number
|
|
summerB += num.Number
|
|
count++
|
|
continue
|
|
}
|
|
num = token.ToBool()
|
|
if num.Type == ArgNumber {
|
|
summerA += num.Number * num.Number
|
|
summerB += num.Number
|
|
count++
|
|
continue
|
|
}
|
|
if name == "VARA" || name == "VARPA" {
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
if count > minimum {
|
|
summerA *= count
|
|
summerB *= summerB
|
|
return newNumberFormulaArg((summerA - summerB) / (count * (count - minimum)))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
|
|
// VAR function returns the sample variance of a supplied set of values. The
|
|
// syntax of the function is:
|
|
//
|
|
// VAR(number1,[number2],...)
|
|
func (fn *formulaFuncs) VAR(argsList *list.List) formulaArg {
|
|
return fn.vars("VAR", argsList)
|
|
}
|
|
|
|
// VARA function calculates the sample variance of a supplied set of values.
|
|
// The syntax of the function is:
|
|
//
|
|
// VARA(number1,[number2],...)
|
|
func (fn *formulaFuncs) VARA(argsList *list.List) formulaArg {
|
|
return fn.vars("VARA", argsList)
|
|
}
|
|
|
|
// VARP function returns the Variance of a given set of values. The syntax of
|
|
// the function is:
|
|
//
|
|
// VARP(number1,[number2],...)
|
|
func (fn *formulaFuncs) VARP(argsList *list.List) formulaArg {
|
|
return fn.vars("VARP", argsList)
|
|
}
|
|
|
|
// VARdotP function returns the Variance of a given set of values. The syntax
|
|
// of the function is:
|
|
//
|
|
// VAR.P(number1,[number2],...)
|
|
func (fn *formulaFuncs) VARdotP(argsList *list.List) formulaArg {
|
|
return fn.vars("VAR.P", argsList)
|
|
}
|
|
|
|
// VARdotS function calculates the sample variance of a supplied set of
|
|
// values. The syntax of the function is:
|
|
//
|
|
// VAR.S(number1,[number2],...)
|
|
func (fn *formulaFuncs) VARdotS(argsList *list.List) formulaArg {
|
|
return fn.vars("VAR.S", argsList)
|
|
}
|
|
|
|
// VARPA function returns the Variance of a given set of values. The syntax of
|
|
// the function is:
|
|
//
|
|
// VARPA(number1,[number2],...)
|
|
func (fn *formulaFuncs) VARPA(argsList *list.List) formulaArg {
|
|
return fn.vars("VARPA", argsList)
|
|
}
|
|
|
|
// WEIBULL function calculates the Weibull Probability Density Function or the
|
|
// Weibull Cumulative Distribution Function for a supplied set of parameters.
|
|
// The syntax of the function is:
|
|
//
|
|
// WEIBULL(x,alpha,beta,cumulative)
|
|
func (fn *formulaFuncs) WEIBULL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEIBULL requires 4 arguments")
|
|
}
|
|
x := argsList.Front().Value.(formulaArg).ToNumber()
|
|
alpha := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
beta := argsList.Back().Prev().Value.(formulaArg).ToNumber()
|
|
if alpha.Type == ArgNumber && beta.Type == ArgNumber && x.Type == ArgNumber {
|
|
if alpha.Number < 0 || alpha.Number <= 0 || beta.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
cumulative := argsList.Back().Value.(formulaArg).ToBool()
|
|
if cumulative.Boolean && cumulative.Number == 1 {
|
|
return newNumberFormulaArg(1 - math.Exp(0-math.Pow(x.Number/beta.Number, alpha.Number)))
|
|
}
|
|
return newNumberFormulaArg((alpha.Number / math.Pow(beta.Number, alpha.Number)) *
|
|
math.Pow(x.Number, alpha.Number-1) * math.Exp(0-math.Pow(x.Number/beta.Number, alpha.Number)))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
|
|
// WEIBULLdotDIST function calculates the Weibull Probability Density Function
|
|
// or the Weibull Cumulative Distribution Function for a supplied set of
|
|
// parameters. The syntax of the function is:
|
|
//
|
|
// WEIBULL.DIST(x,alpha,beta,cumulative)
|
|
func (fn *formulaFuncs) WEIBULLdotDIST(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEIBULL.DIST requires 4 arguments")
|
|
}
|
|
return fn.WEIBULL(argsList)
|
|
}
|
|
|
|
// ZdotTEST function calculates the one-tailed probability value of the
|
|
// Z-Test. The syntax of the function is:
|
|
//
|
|
// Z.TEST(array,x,[sigma])
|
|
func (fn *formulaFuncs) ZdotTEST(argsList *list.List) formulaArg {
|
|
argsLen := argsList.Len()
|
|
if argsLen < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "Z.TEST requires at least 2 arguments")
|
|
}
|
|
if argsLen > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "Z.TEST accepts at most 3 arguments")
|
|
}
|
|
return fn.ZTEST(argsList)
|
|
}
|
|
|
|
// ZTEST function calculates the one-tailed probability value of the Z-Test.
|
|
// The syntax of the function is:
|
|
//
|
|
// ZTEST(array,x,[sigma])
|
|
func (fn *formulaFuncs) ZTEST(argsList *list.List) formulaArg {
|
|
argsLen := argsList.Len()
|
|
if argsLen < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ZTEST requires at least 2 arguments")
|
|
}
|
|
if argsLen > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ZTEST accepts at most 3 arguments")
|
|
}
|
|
arrArg, arrArgs := argsList.Front().Value.(formulaArg), list.New()
|
|
arrArgs.PushBack(arrArg)
|
|
arr := fn.AVERAGE(arrArgs)
|
|
if arr.Type == ArgError {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
x := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if x.Type == ArgError {
|
|
return x
|
|
}
|
|
sigma := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if sigma.Type == ArgError {
|
|
return sigma
|
|
}
|
|
if argsLen != 3 {
|
|
sigma = fn.STDEV(arrArgs).ToNumber()
|
|
}
|
|
normsdistArg := list.New()
|
|
div := sigma.Number / math.Sqrt(float64(len(arrArg.ToList())))
|
|
if div == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
normsdistArg.PushBack(newNumberFormulaArg((arr.Number - x.Number) / div))
|
|
return newNumberFormulaArg(1 - fn.NORMSDIST(normsdistArg).Number)
|
|
}
|
|
|
|
// Information Functions
|
|
|
|
// ERRORdotTYPE function receives an error value and returns an integer, that
|
|
// tells you the type of the supplied error. The syntax of the function is:
|
|
//
|
|
// ERROR.TYPE(error_val)
|
|
func (fn *formulaFuncs) ERRORdotTYPE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ERROR.TYPE requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
if token.Type == ArgError {
|
|
for i, errType := range []string{
|
|
formulaErrorNULL, formulaErrorDIV, formulaErrorVALUE, formulaErrorREF,
|
|
formulaErrorNAME, formulaErrorNUM, formulaErrorNA,
|
|
} {
|
|
if errType == token.String {
|
|
return newNumberFormulaArg(float64(i) + 1)
|
|
}
|
|
}
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// ISBLANK function tests if a specified cell is blank (empty) and if so,
|
|
// returns TRUE; Otherwise the function returns FALSE. The syntax of the
|
|
// function is:
|
|
//
|
|
// ISBLANK(value)
|
|
func (fn *formulaFuncs) ISBLANK(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISBLANK requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgUnknown, ArgEmpty:
|
|
return newBoolFormulaArg(true)
|
|
default:
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
}
|
|
|
|
// ISERR function tests if an initial supplied expression (or value) returns
|
|
// any Excel Error, except the #N/A error. If so, the function returns the
|
|
// logical value TRUE; If the supplied value is not an error or is the #N/A
|
|
// error, the ISERR function returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISERR(value)
|
|
func (fn *formulaFuncs) ISERR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISERR requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
result := false
|
|
if token.Type == ArgError {
|
|
for _, errType := range []string{
|
|
formulaErrorDIV, formulaErrorNAME, formulaErrorNUM,
|
|
formulaErrorVALUE, formulaErrorREF, formulaErrorNULL,
|
|
formulaErrorSPILL, formulaErrorCALC, formulaErrorGETTINGDATA,
|
|
} {
|
|
if errType == token.String {
|
|
result = true
|
|
}
|
|
}
|
|
}
|
|
return newBoolFormulaArg(result)
|
|
}
|
|
|
|
// ISERROR function tests if an initial supplied expression (or value) returns
|
|
// an Excel Error, and if so, returns the logical value TRUE; Otherwise the
|
|
// function returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISERROR(value)
|
|
func (fn *formulaFuncs) ISERROR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISERROR requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
result := false
|
|
if token.Type == ArgError {
|
|
for _, errType := range []string{
|
|
formulaErrorDIV, formulaErrorNAME, formulaErrorNA, formulaErrorNUM,
|
|
formulaErrorVALUE, formulaErrorREF, formulaErrorNULL, formulaErrorSPILL,
|
|
formulaErrorCALC, formulaErrorGETTINGDATA,
|
|
} {
|
|
if errType == token.String {
|
|
result = true
|
|
}
|
|
}
|
|
}
|
|
return newBoolFormulaArg(result)
|
|
}
|
|
|
|
// ISEVEN function tests if a supplied number (or numeric expression)
|
|
// evaluates to an even number, and if so, returns TRUE; Otherwise, the
|
|
// function returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISEVEN(value)
|
|
func (fn *formulaFuncs) ISEVEN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISEVEN requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgEmpty:
|
|
return newBoolFormulaArg(true)
|
|
case ArgNumber, ArgString:
|
|
num := token.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if num.Number == 1 {
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
return newBoolFormulaArg(num.Number == num.Number/2*2)
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
|
|
// ISFORMULA function tests if a specified cell contains a formula, and if so,
|
|
// returns TRUE; Otherwise, the function returns FALSE. The syntax of the
|
|
// function is:
|
|
//
|
|
// ISFORMULA(reference)
|
|
func (fn *formulaFuncs) ISFORMULA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISFORMULA requires 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
if arg.cellRefs != nil && arg.cellRefs.Len() == 1 {
|
|
ref := arg.cellRefs.Front().Value.(cellRef)
|
|
cell, _ := CoordinatesToCellName(ref.Col, ref.Row)
|
|
if formula, _ := fn.f.GetCellFormula(ref.Sheet, cell); len(formula) > 0 {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// ISLOGICAL function tests if a supplied value (or expression) returns a
|
|
// logical value (i.e. evaluates to True or False). If so, the function
|
|
// returns TRUE; Otherwise, it returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISLOGICAL(value)
|
|
func (fn *formulaFuncs) ISLOGICAL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISLOGICAL requires 1 argument")
|
|
}
|
|
val := argsList.Front().Value.(formulaArg).Value()
|
|
if strings.EqualFold("TRUE", val) || strings.EqualFold("FALSE", val) {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// ISNA function tests if an initial supplied expression (or value) returns
|
|
// the Excel #N/A Error, and if so, returns TRUE; Otherwise the function
|
|
// returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISNA(value)
|
|
func (fn *formulaFuncs) ISNA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISNA requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
result := "FALSE"
|
|
if token.Type == ArgError && token.String == formulaErrorNA {
|
|
result = "TRUE"
|
|
}
|
|
return newStringFormulaArg(result)
|
|
}
|
|
|
|
// ISNONTEXT function tests if a supplied value is text. If not, the
|
|
// function returns TRUE; If the supplied value is text, the function returns
|
|
// FALSE. The syntax of the function is:
|
|
//
|
|
// ISNONTEXT(value)
|
|
func (fn *formulaFuncs) ISNONTEXT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISNONTEXT requires 1 argument")
|
|
}
|
|
if argsList.Front().Value.(formulaArg).Type == ArgString {
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
|
|
// ISNUMBER function tests if a supplied value is a number. If so,
|
|
// the function returns TRUE; Otherwise it returns FALSE. The syntax of the
|
|
// function is:
|
|
//
|
|
// ISNUMBER(value)
|
|
func (fn *formulaFuncs) ISNUMBER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISNUMBER requires 1 argument")
|
|
}
|
|
if argsList.Front().Value.(formulaArg).Type == ArgNumber {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// ISODD function tests if a supplied number (or numeric expression) evaluates
|
|
// to an odd number, and if so, returns TRUE; Otherwise, the function returns
|
|
// FALSE. The syntax of the function is:
|
|
//
|
|
// ISODD(value)
|
|
func (fn *formulaFuncs) ISODD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISODD requires 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if int(arg.Number) != int(arg.Number)/2*2 {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// ISREF function tests if a supplied value is a reference. If so, the
|
|
// function returns TRUE; Otherwise it returns FALSE. The syntax of the
|
|
// function is:
|
|
//
|
|
// ISREF(value)
|
|
func (fn *formulaFuncs) ISREF(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISREF requires 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
if arg.cellRanges != nil && arg.cellRanges.Len() > 0 || arg.cellRefs != nil && arg.cellRefs.Len() > 0 {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// ISTEXT function tests if a supplied value is text, and if so, returns TRUE;
|
|
// Otherwise, the function returns FALSE. The syntax of the function is:
|
|
//
|
|
// ISTEXT(value)
|
|
func (fn *formulaFuncs) ISTEXT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISTEXT requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
if token.ToNumber().Type != ArgError {
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
return newBoolFormulaArg(token.Type == ArgString)
|
|
}
|
|
|
|
// N function converts data into a numeric value. The syntax of the function
|
|
// is:
|
|
//
|
|
// N(value)
|
|
func (fn *formulaFuncs) N(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "N requires 1 argument")
|
|
}
|
|
token, num := argsList.Front().Value.(formulaArg), 0.0
|
|
if token.Type == ArgError {
|
|
return token
|
|
}
|
|
if arg := token.ToNumber(); arg.Type == ArgNumber {
|
|
num = arg.Number
|
|
}
|
|
if token.Value() == "TRUE" {
|
|
num = 1
|
|
}
|
|
return newNumberFormulaArg(num)
|
|
}
|
|
|
|
// NA function returns the Excel #N/A error. This error message has the
|
|
// meaning 'value not available' and is produced when an Excel Formula is
|
|
// unable to find a value that it needs. The syntax of the function is:
|
|
//
|
|
// NA()
|
|
func (fn *formulaFuncs) NA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NA accepts no arguments")
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// SHEET function returns the Sheet number for a specified reference. The
|
|
// syntax of the function is:
|
|
//
|
|
// SHEET([value])
|
|
func (fn *formulaFuncs) SHEET(argsList *list.List) formulaArg {
|
|
if argsList.Len() > 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SHEET accepts at most 1 argument")
|
|
}
|
|
if argsList.Len() == 0 {
|
|
idx, _ := fn.f.GetSheetIndex(fn.sheet)
|
|
return newNumberFormulaArg(float64(idx + 1))
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
if sheetIdx, _ := fn.f.GetSheetIndex(arg.Value()); sheetIdx != -1 {
|
|
return newNumberFormulaArg(float64(sheetIdx + 1))
|
|
}
|
|
if arg.cellRanges != nil && arg.cellRanges.Len() > 0 {
|
|
if sheetIdx, _ := fn.f.GetSheetIndex(arg.cellRanges.Front().Value.(cellRange).From.Sheet); sheetIdx != -1 {
|
|
return newNumberFormulaArg(float64(sheetIdx + 1))
|
|
}
|
|
}
|
|
if arg.cellRefs != nil && arg.cellRefs.Len() > 0 {
|
|
if sheetIdx, _ := fn.f.GetSheetIndex(arg.cellRefs.Front().Value.(cellRef).Sheet); sheetIdx != -1 {
|
|
return newNumberFormulaArg(float64(sheetIdx + 1))
|
|
}
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// SHEETS function returns the number of sheets in a supplied reference. The
|
|
// result includes sheets that are Visible, Hidden or Very Hidden. The syntax
|
|
// of the function is:
|
|
//
|
|
// SHEETS([reference])
|
|
func (fn *formulaFuncs) SHEETS(argsList *list.List) formulaArg {
|
|
if argsList.Len() > 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SHEETS accepts at most 1 argument")
|
|
}
|
|
if argsList.Len() == 0 {
|
|
return newNumberFormulaArg(float64(len(fn.f.GetSheetList())))
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
sheetMap := map[string]struct{}{}
|
|
if arg.cellRanges != nil && arg.cellRanges.Len() > 0 {
|
|
for rng := arg.cellRanges.Front(); rng != nil; rng = rng.Next() {
|
|
sheetMap[rng.Value.(cellRange).From.Sheet] = struct{}{}
|
|
}
|
|
}
|
|
if arg.cellRefs != nil && arg.cellRefs.Len() > 0 {
|
|
for ref := arg.cellRefs.Front(); ref != nil; ref = ref.Next() {
|
|
sheetMap[ref.Value.(cellRef).Sheet] = struct{}{}
|
|
}
|
|
}
|
|
if len(sheetMap) > 0 {
|
|
return newNumberFormulaArg(float64(len(sheetMap)))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// TYPE function returns an integer that represents the value's data type. The
|
|
// syntax of the function is:
|
|
//
|
|
// TYPE(value)
|
|
func (fn *formulaFuncs) TYPE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TYPE requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgError:
|
|
return newNumberFormulaArg(16)
|
|
case ArgMatrix:
|
|
return newNumberFormulaArg(64)
|
|
case ArgNumber, ArgEmpty:
|
|
if token.Boolean {
|
|
return newNumberFormulaArg(4)
|
|
}
|
|
return newNumberFormulaArg(1)
|
|
default:
|
|
return newNumberFormulaArg(2)
|
|
}
|
|
}
|
|
|
|
// T function tests if a supplied value is text and if so, returns the
|
|
// supplied text; Otherwise, the function returns an empty text string. The
|
|
// syntax of the function is:
|
|
//
|
|
// T(value)
|
|
func (fn *formulaFuncs) T(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "T requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
if token.Type == ArgError {
|
|
return token
|
|
}
|
|
if token.Type == ArgNumber {
|
|
return newStringFormulaArg("")
|
|
}
|
|
return newStringFormulaArg(token.Value())
|
|
}
|
|
|
|
// Logical Functions
|
|
|
|
// AND function tests a number of supplied conditions and returns TRUE or
|
|
// FALSE. The syntax of the function is:
|
|
//
|
|
// AND(logical_test1,[logical_test2],...)
|
|
func (fn *formulaFuncs) AND(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AND requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 30 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AND accepts at most 30 arguments")
|
|
}
|
|
and := true
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgUnknown:
|
|
continue
|
|
case ArgString:
|
|
if token.String == "TRUE" {
|
|
continue
|
|
}
|
|
if token.String == "FALSE" {
|
|
return newStringFormulaArg(token.String)
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
case ArgNumber:
|
|
and = and && token.Number != 0
|
|
case ArgMatrix:
|
|
// TODO
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
return newBoolFormulaArg(and)
|
|
}
|
|
|
|
// FALSE function returns the logical value FALSE. The syntax of the
|
|
// function is:
|
|
//
|
|
// FALSE()
|
|
func (fn *formulaFuncs) FALSE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FALSE takes no arguments")
|
|
}
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
|
|
// IFERROR function receives two values (or expressions) and tests if the
|
|
// first of these evaluates to an error. The syntax of the function is:
|
|
//
|
|
// IFERROR(value,value_if_error)
|
|
func (fn *formulaFuncs) IFERROR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IFERROR requires 2 arguments")
|
|
}
|
|
value := argsList.Front().Value.(formulaArg)
|
|
if value.Type != ArgError {
|
|
if value.Type == ArgEmpty {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
return value
|
|
}
|
|
return argsList.Back().Value.(formulaArg)
|
|
}
|
|
|
|
// IFNA function tests if an initial supplied value (or expression) evaluates
|
|
// to the Excel #N/A error. If so, the function returns a second supplied
|
|
// value; Otherwise the function returns the first supplied value. The syntax
|
|
// of the function is:
|
|
//
|
|
// IFNA(value,value_if_na)
|
|
func (fn *formulaFuncs) IFNA(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IFNA requires 2 arguments")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
if arg.Type == ArgError && arg.Value() == formulaErrorNA {
|
|
return argsList.Back().Value.(formulaArg)
|
|
}
|
|
return arg
|
|
}
|
|
|
|
// IFS function tests a number of supplied conditions and returns the result
|
|
// corresponding to the first condition that evaluates to TRUE. If none of
|
|
// the supplied conditions evaluate to TRUE, the function returns the #N/A
|
|
// error.
|
|
//
|
|
// IFS(logical_test1,value_if_true1,[logical_test2,value_if_true2],...)
|
|
func (fn *formulaFuncs) IFS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IFS requires at least 2 arguments")
|
|
}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
if arg.Value.(formulaArg).ToBool().Number == 1 {
|
|
return arg.Next().Value.(formulaArg)
|
|
}
|
|
arg = arg.Next()
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// NOT function returns the opposite to a supplied logical value. The syntax
|
|
// of the function is:
|
|
//
|
|
// NOT(logical)
|
|
func (fn *formulaFuncs) NOT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NOT requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString, ArgList:
|
|
if strings.ToUpper(token.String) == "TRUE" {
|
|
return newBoolFormulaArg(false)
|
|
}
|
|
if strings.ToUpper(token.String) == "FALSE" {
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
case ArgNumber:
|
|
return newBoolFormulaArg(!(token.Number != 0))
|
|
case ArgError:
|
|
return token
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NOT expects 1 boolean or numeric argument")
|
|
}
|
|
|
|
// OR function tests a number of supplied conditions and returns either TRUE
|
|
// or FALSE. The syntax of the function is:
|
|
//
|
|
// OR(logical_test1,[logical_test2],...)
|
|
func (fn *formulaFuncs) OR(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OR requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 30 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "OR accepts at most 30 arguments")
|
|
}
|
|
var or bool
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgUnknown:
|
|
continue
|
|
case ArgString:
|
|
if token.String == "FALSE" {
|
|
continue
|
|
}
|
|
if token.String == "TRUE" {
|
|
or = true
|
|
continue
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
case ArgNumber:
|
|
if or = token.Number != 0; or {
|
|
return newStringFormulaArg(strings.ToUpper(strconv.FormatBool(or)))
|
|
}
|
|
case ArgMatrix:
|
|
// TODO
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(strconv.FormatBool(or)))
|
|
}
|
|
|
|
// SWITCH function compares a number of supplied values to a supplied test
|
|
// expression and returns a result corresponding to the first value that
|
|
// matches the test expression. A default value can be supplied, to be
|
|
// returned if none of the supplied values match the test expression. The
|
|
// syntax of the function is:
|
|
//
|
|
// SWITCH(expression,value1,result1,[value2,result2],[value3,result3],...,[default])
|
|
func (fn *formulaFuncs) SWITCH(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SWITCH requires at least 3 arguments")
|
|
}
|
|
target := argsList.Front().Value.(formulaArg)
|
|
argCount := argsList.Len() - 1
|
|
switchCount := int(math.Floor(float64(argCount) / 2))
|
|
hasDefaultClause := argCount%2 != 0
|
|
result := newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
if hasDefaultClause {
|
|
result = argsList.Back().Value.(formulaArg)
|
|
}
|
|
if switchCount > 0 {
|
|
arg := argsList.Front()
|
|
for i := 0; i < switchCount; i++ {
|
|
arg = arg.Next()
|
|
if target.Value() == arg.Value.(formulaArg).Value() {
|
|
result = arg.Next().Value.(formulaArg)
|
|
break
|
|
}
|
|
arg = arg.Next()
|
|
}
|
|
}
|
|
return result
|
|
}
|
|
|
|
// TRUE function returns the logical value TRUE. The syntax of the function
|
|
// is:
|
|
//
|
|
// TRUE()
|
|
func (fn *formulaFuncs) TRUE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TRUE takes no arguments")
|
|
}
|
|
return newBoolFormulaArg(true)
|
|
}
|
|
|
|
// calcXor checking if numeric cell exists and count it by given arguments
|
|
// sequence for the formula function XOR.
|
|
func calcXor(argsList *list.List) formulaArg {
|
|
count, ok := 0, false
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgError:
|
|
return token
|
|
case ArgNumber:
|
|
ok = true
|
|
if token.Number != 0 {
|
|
count++
|
|
}
|
|
case ArgMatrix:
|
|
for _, value := range token.ToList() {
|
|
if num := value.ToNumber(); num.Type == ArgNumber {
|
|
ok = true
|
|
if num.Number != 0 {
|
|
count++
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newBoolFormulaArg(count%2 != 0)
|
|
}
|
|
|
|
// XOR function returns the Exclusive Or logical operation for one or more
|
|
// supplied conditions. I.e. the Xor function returns TRUE if an odd number
|
|
// of the supplied conditions evaluate to TRUE, and FALSE otherwise. The
|
|
// syntax of the function is:
|
|
//
|
|
// XOR(logical_test1,[logical_test2],...)
|
|
func (fn *formulaFuncs) XOR(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XOR requires at least 1 argument")
|
|
}
|
|
return calcXor(argsList)
|
|
}
|
|
|
|
// Date and Time Functions
|
|
|
|
// DATE returns a date, from a user-supplied year, month and day. The syntax
|
|
// of the function is:
|
|
//
|
|
// DATE(year,month,day)
|
|
func (fn *formulaFuncs) DATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DATE requires 3 number arguments")
|
|
}
|
|
year := argsList.Front().Value.(formulaArg).ToNumber()
|
|
month := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
day := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if year.Type != ArgNumber || month.Type != ArgNumber || day.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DATE requires 3 number arguments")
|
|
}
|
|
d := makeDate(int(year.Number), time.Month(month.Number), int(day.Number))
|
|
return newStringFormulaArg(timeFromExcelTime(daysBetween(excelMinTime1900.Unix(), d)+1, false).String())
|
|
}
|
|
|
|
// calcDateDif is an implementation of the formula function DATEDIF,
|
|
// calculation difference between two dates.
|
|
func calcDateDif(unit string, diff float64, seq []int, startArg, endArg formulaArg) float64 {
|
|
ey, sy, em, sm, ed, sd := seq[0], seq[1], seq[2], seq[3], seq[4], seq[5]
|
|
switch unit {
|
|
case "d":
|
|
diff = endArg.Number - startArg.Number
|
|
case "md":
|
|
smMD := em
|
|
if ed < sd {
|
|
smMD--
|
|
}
|
|
diff = endArg.Number - daysBetween(excelMinTime1900.Unix(), makeDate(ey, time.Month(smMD), sd)) - 1
|
|
case "ym":
|
|
diff = float64(em - sm)
|
|
if ed < sd {
|
|
diff--
|
|
}
|
|
if diff < 0 {
|
|
diff += 12
|
|
}
|
|
case "yd":
|
|
syYD := sy
|
|
if em < sm || (em == sm && ed < sd) {
|
|
syYD++
|
|
}
|
|
s := daysBetween(excelMinTime1900.Unix(), makeDate(syYD, time.Month(em), ed))
|
|
e := daysBetween(excelMinTime1900.Unix(), makeDate(sy, time.Month(sm), sd))
|
|
diff = s - e
|
|
}
|
|
return diff
|
|
}
|
|
|
|
// DATEDIF function calculates the number of days, months, or years between
|
|
// two dates. The syntax of the function is:
|
|
//
|
|
// DATEDIF(start_date,end_date,unit)
|
|
func (fn *formulaFuncs) DATEDIF(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DATEDIF requires 3 number arguments")
|
|
}
|
|
startArg, endArg := argsList.Front().Value.(formulaArg).ToNumber(), argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if startArg.Type != ArgNumber || endArg.Type != ArgNumber {
|
|
return startArg
|
|
}
|
|
if startArg.Number > endArg.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "start_date > end_date")
|
|
}
|
|
if startArg.Number == endArg.Number {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
unit := strings.ToLower(argsList.Back().Value.(formulaArg).Value())
|
|
startDate, endDate := timeFromExcelTime(startArg.Number, false), timeFromExcelTime(endArg.Number, false)
|
|
sy, smm, sd := startDate.Date()
|
|
ey, emm, ed := endDate.Date()
|
|
sm, em, diff := int(smm), int(emm), 0.0
|
|
switch unit {
|
|
case "y":
|
|
diff = float64(ey - sy)
|
|
if em < sm || (em == sm && ed < sd) {
|
|
diff--
|
|
}
|
|
case "m":
|
|
yDiff := ey - sy
|
|
mDiff := em - sm
|
|
if ed < sd {
|
|
mDiff--
|
|
}
|
|
if mDiff < 0 {
|
|
yDiff--
|
|
mDiff += 12
|
|
}
|
|
diff = float64(yDiff*12 + mDiff)
|
|
case "d", "md", "ym", "yd":
|
|
diff = calcDateDif(unit, diff, []int{ey, sy, em, sm, ed, sd}, startArg, endArg)
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DATEDIF has invalid unit")
|
|
}
|
|
return newNumberFormulaArg(diff)
|
|
}
|
|
|
|
// isDateOnlyFmt check if the given string matches date-only format regular expressions.
|
|
func isDateOnlyFmt(dateString string) bool {
|
|
for _, df := range dateOnlyFormats {
|
|
subMatch := df.FindStringSubmatch(dateString)
|
|
if len(subMatch) > 1 {
|
|
return true
|
|
}
|
|
}
|
|
return false
|
|
}
|
|
|
|
// isTimeOnlyFmt check if the given string matches time-only format regular expressions.
|
|
func isTimeOnlyFmt(timeString string) bool {
|
|
for _, tf := range timeFormats {
|
|
subMatch := tf.FindStringSubmatch(timeString)
|
|
if len(subMatch) > 1 {
|
|
return true
|
|
}
|
|
}
|
|
return false
|
|
}
|
|
|
|
// strToTimePatternHandler1 parse and convert the given string in pattern
|
|
// hh to the time.
|
|
func strToTimePatternHandler1(subMatch []string) (h, m int, s float64, err error) {
|
|
h, err = strconv.Atoi(subMatch[0])
|
|
return
|
|
}
|
|
|
|
// strToTimePatternHandler2 parse and convert the given string in pattern
|
|
// hh:mm to the time.
|
|
func strToTimePatternHandler2(subMatch []string) (h, m int, s float64, err error) {
|
|
if h, err = strconv.Atoi(subMatch[0]); err != nil {
|
|
return
|
|
}
|
|
m, err = strconv.Atoi(subMatch[2])
|
|
return
|
|
}
|
|
|
|
// strToTimePatternHandler3 parse and convert the given string in pattern
|
|
// mm:ss to the time.
|
|
func strToTimePatternHandler3(subMatch []string) (h, m int, s float64, err error) {
|
|
if m, err = strconv.Atoi(subMatch[0]); err != nil {
|
|
return
|
|
}
|
|
s, err = strconv.ParseFloat(subMatch[2], 64)
|
|
return
|
|
}
|
|
|
|
// strToTimePatternHandler4 parse and convert the given string in pattern
|
|
// hh:mm:ss to the time.
|
|
func strToTimePatternHandler4(subMatch []string) (h, m int, s float64, err error) {
|
|
if h, err = strconv.Atoi(subMatch[0]); err != nil {
|
|
return
|
|
}
|
|
if m, err = strconv.Atoi(subMatch[2]); err != nil {
|
|
return
|
|
}
|
|
s, err = strconv.ParseFloat(subMatch[4], 64)
|
|
return
|
|
}
|
|
|
|
// strToTime parse and convert the given string to the time.
|
|
func strToTime(str string) (int, int, float64, bool, bool, formulaArg) {
|
|
var subMatch []string
|
|
pattern := ""
|
|
for key, tf := range timeFormats {
|
|
subMatch = tf.FindStringSubmatch(str)
|
|
if len(subMatch) > 1 {
|
|
pattern = key
|
|
break
|
|
}
|
|
}
|
|
if pattern == "" {
|
|
return 0, 0, 0, false, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
dateIsEmpty := subMatch[1] == ""
|
|
subMatch = subMatch[49:]
|
|
var (
|
|
l = len(subMatch)
|
|
last = subMatch[l-1]
|
|
am = last == "am"
|
|
pm = last == "pm"
|
|
hours, minutes int
|
|
seconds float64
|
|
err error
|
|
)
|
|
if handler, ok := map[string]func(match []string) (int, int, float64, error){
|
|
"hh": strToTimePatternHandler1,
|
|
"hh:mm": strToTimePatternHandler2,
|
|
"mm:ss": strToTimePatternHandler3,
|
|
"hh:mm:ss": strToTimePatternHandler4,
|
|
}[pattern]; ok {
|
|
if hours, minutes, seconds, err = handler(subMatch); err != nil {
|
|
return 0, 0, 0, false, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
if minutes >= 60 {
|
|
return 0, 0, 0, false, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if am || pm {
|
|
if hours > 12 || seconds >= 60 {
|
|
return 0, 0, 0, false, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
} else if hours == 12 {
|
|
hours = 0
|
|
}
|
|
} else if hours >= 24 || seconds >= 10000 {
|
|
return 0, 0, 0, false, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return hours, minutes, seconds, pm, dateIsEmpty, newEmptyFormulaArg()
|
|
}
|
|
|
|
// strToDatePatternHandler1 parse and convert the given string in pattern
|
|
// mm/dd/yy to the date.
|
|
func strToDatePatternHandler1(subMatch []string) (int, int, int, bool, error) {
|
|
var year, month, day int
|
|
var err error
|
|
if month, err = strconv.Atoi(subMatch[1]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if day, err = strconv.Atoi(subMatch[3]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if year, err = strconv.Atoi(subMatch[5]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if year < 0 || year > 9999 || (year > 99 && year < 1900) {
|
|
return 0, 0, 0, false, ErrParameterInvalid
|
|
}
|
|
return formatYear(year), month, day, subMatch[8] == "", err
|
|
}
|
|
|
|
// strToDatePatternHandler2 parse and convert the given string in pattern mm
|
|
// dd, yy to the date.
|
|
func strToDatePatternHandler2(subMatch []string) (int, int, int, bool, error) {
|
|
var year, month, day int
|
|
var err error
|
|
month = month2num[subMatch[1]]
|
|
if day, err = strconv.Atoi(subMatch[14]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if year, err = strconv.Atoi(subMatch[16]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if year < 0 || year > 9999 || (year > 99 && year < 1900) {
|
|
return 0, 0, 0, false, ErrParameterInvalid
|
|
}
|
|
return formatYear(year), month, day, subMatch[19] == "", err
|
|
}
|
|
|
|
// strToDatePatternHandler3 parse and convert the given string in pattern
|
|
// yy-mm-dd to the date.
|
|
func strToDatePatternHandler3(subMatch []string) (int, int, int, bool, error) {
|
|
var year, month, day int
|
|
v1, err := strconv.Atoi(subMatch[1])
|
|
if err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
v2, err := strconv.Atoi(subMatch[3])
|
|
if err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
v3, err := strconv.Atoi(subMatch[5])
|
|
if err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
if v1 >= 1900 && v1 < 10000 {
|
|
year = v1
|
|
month = v2
|
|
day = v3
|
|
} else if v1 > 0 && v1 < 13 {
|
|
month = v1
|
|
day = v2
|
|
year = v3
|
|
} else {
|
|
return 0, 0, 0, false, ErrParameterInvalid
|
|
}
|
|
return year, month, day, subMatch[8] == "", err
|
|
}
|
|
|
|
// strToDatePatternHandler4 parse and convert the given string in pattern
|
|
// yy-mmStr-dd, yy to the date.
|
|
func strToDatePatternHandler4(subMatch []string) (int, int, int, bool, error) {
|
|
var year, month, day int
|
|
var err error
|
|
if year, err = strconv.Atoi(subMatch[16]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
month = month2num[subMatch[3]]
|
|
if day, err = strconv.Atoi(subMatch[1]); err != nil {
|
|
return 0, 0, 0, false, err
|
|
}
|
|
return formatYear(year), month, day, subMatch[19] == "", err
|
|
}
|
|
|
|
// strToDate parse and convert the given string to the date.
|
|
func strToDate(str string) (int, int, int, bool, formulaArg) {
|
|
var subMatch []string
|
|
pattern := ""
|
|
for key, df := range dateFormats {
|
|
subMatch = df.FindStringSubmatch(str)
|
|
if len(subMatch) > 1 {
|
|
pattern = key
|
|
break
|
|
}
|
|
}
|
|
if pattern == "" {
|
|
return 0, 0, 0, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
var (
|
|
timeIsEmpty bool
|
|
year, month, day int
|
|
err error
|
|
)
|
|
if handler, ok := map[string]func(match []string) (int, int, int, bool, error){
|
|
"mm/dd/yy": strToDatePatternHandler1,
|
|
"mm dd, yy": strToDatePatternHandler2,
|
|
"yy-mm-dd": strToDatePatternHandler3,
|
|
"yy-mmStr-dd": strToDatePatternHandler4,
|
|
}[pattern]; ok {
|
|
if year, month, day, timeIsEmpty, err = handler(subMatch); err != nil {
|
|
return 0, 0, 0, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
if !validateDate(year, month, day) {
|
|
return 0, 0, 0, false, newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return year, month, day, timeIsEmpty, newEmptyFormulaArg()
|
|
}
|
|
|
|
// DATEVALUE function converts a text representation of a date into an Excel
|
|
// date. For example, the function converts a text string representing a
|
|
// date, into the serial number that represents the date in Excels' date-time
|
|
// code. The syntax of the function is:
|
|
//
|
|
// DATEVALUE(date_text)
|
|
func (fn *formulaFuncs) DATEVALUE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DATEVALUE requires 1 argument")
|
|
}
|
|
dateText := argsList.Front().Value.(formulaArg).Value()
|
|
if !isDateOnlyFmt(dateText) {
|
|
if _, _, _, _, _, err := strToTime(dateText); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateText)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(daysBetween(excelMinTime1900.Unix(), makeDate(y, time.Month(m), d)) + 1)
|
|
}
|
|
|
|
// DAY function returns the day of a date, represented by a serial number. The
|
|
// day is given as an integer ranging from 1 to 31. The syntax of the
|
|
// function is:
|
|
//
|
|
// DAY(serial_number)
|
|
func (fn *formulaFuncs) DAY(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DAY requires exactly 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(arg.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
_, _, day, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(float64(day))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "DAY only accepts positive argument")
|
|
}
|
|
if num.Number <= 60 {
|
|
return newNumberFormulaArg(math.Mod(num.Number, 31.0))
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Day()))
|
|
}
|
|
|
|
// DAYS function returns the number of days between two supplied dates. The
|
|
// syntax of the function is:
|
|
//
|
|
// DAYS(end_date,start_date)
|
|
func (fn *formulaFuncs) DAYS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DAYS requires 2 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
end, start := args.List[0], args.List[1]
|
|
return newNumberFormulaArg(end.Number - start.Number)
|
|
}
|
|
|
|
// DAYS360 function returns the number of days between 2 dates, based on a
|
|
// 360-day year (12 x 30 months). The syntax of the function is:
|
|
//
|
|
// DAYS360(start_date,end_date,[method])
|
|
func (fn *formulaFuncs) DAYS360(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DAYS360 requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DAYS360 requires at most 3 arguments")
|
|
}
|
|
startDate := toExcelDateArg(argsList.Front().Value.(formulaArg))
|
|
if startDate.Type != ArgNumber {
|
|
return startDate
|
|
}
|
|
endDate := toExcelDateArg(argsList.Front().Next().Value.(formulaArg))
|
|
if endDate.Type != ArgNumber {
|
|
return endDate
|
|
}
|
|
start, end := timeFromExcelTime(startDate.Number, false), timeFromExcelTime(endDate.Number, false)
|
|
sy, sm, sd, ey, em, ed := start.Year(), int(start.Month()), start.Day(), end.Year(), int(end.Month()), end.Day()
|
|
method := newBoolFormulaArg(false)
|
|
if argsList.Len() > 2 {
|
|
if method = argsList.Back().Value.(formulaArg).ToBool(); method.Type != ArgNumber {
|
|
return method
|
|
}
|
|
}
|
|
if method.Number == 1 {
|
|
if sd == 31 {
|
|
sd--
|
|
}
|
|
if ed == 31 {
|
|
ed--
|
|
}
|
|
} else {
|
|
if getDaysInMonth(sy, sm) == sd {
|
|
sd = 30
|
|
}
|
|
if ed > 30 {
|
|
if sd < 30 {
|
|
em++
|
|
ed = 1
|
|
} else {
|
|
ed = 30
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(float64(360*(ey-sy) + 30*(em-sm) + (ed - sd)))
|
|
}
|
|
|
|
// ISOWEEKNUM function returns the ISO week number of a supplied date. The
|
|
// syntax of the function is:
|
|
//
|
|
// ISOWEEKNUM(date)
|
|
func (fn *formulaFuncs) ISOWEEKNUM(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISOWEEKNUM requires 1 argument")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
weekNum := 0
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(date.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
_, weekNum = time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.UTC).ISOWeek()
|
|
} else {
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
_, weekNum = timeFromExcelTime(num.Number, false).ISOWeek()
|
|
}
|
|
return newNumberFormulaArg(float64(weekNum))
|
|
}
|
|
|
|
// EDATE function returns a date that is a specified number of months before or
|
|
// after a supplied start date. The syntax of function is:
|
|
//
|
|
// EDATE(start_date,months)
|
|
func (fn *formulaFuncs) EDATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EDATE requires 2 arguments")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
var dateTime time.Time
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(date.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
dateTime = time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.Now().Location())
|
|
} else {
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
dateTime = timeFromExcelTime(num.Number, false)
|
|
}
|
|
month := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if month.Type != ArgNumber {
|
|
return month
|
|
}
|
|
y, d := dateTime.Year(), dateTime.Day()
|
|
m := int(dateTime.Month()) + int(month.Number)
|
|
if month.Number < 0 {
|
|
y -= int(math.Ceil(-1 * float64(m) / 12))
|
|
}
|
|
if month.Number > 11 {
|
|
y += int(math.Floor(float64(m) / 12))
|
|
}
|
|
if m = m % 12; m < 0 {
|
|
m += 12
|
|
}
|
|
if d > 28 {
|
|
if days := getDaysInMonth(y, m); d > days {
|
|
d = days
|
|
}
|
|
}
|
|
result, _ := timeToExcelTime(time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.UTC), false)
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// EOMONTH function returns the last day of the month, that is a specified
|
|
// number of months before or after an initial supplied start date. The syntax
|
|
// of the function is:
|
|
//
|
|
// EOMONTH(start_date,months)
|
|
func (fn *formulaFuncs) EOMONTH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EOMONTH requires 2 arguments")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
var dateTime time.Time
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(date.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
dateTime = time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.Now().Location())
|
|
} else {
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
dateTime = timeFromExcelTime(num.Number, false)
|
|
}
|
|
months := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if months.Type != ArgNumber {
|
|
return months
|
|
}
|
|
y, m := dateTime.Year(), int(dateTime.Month())+int(months.Number)-1
|
|
if m < 0 {
|
|
y -= int(math.Ceil(-1 * float64(m) / 12))
|
|
}
|
|
if m > 11 {
|
|
y += int(math.Floor(float64(m) / 12))
|
|
}
|
|
if m = m % 12; m < 0 {
|
|
m += 12
|
|
}
|
|
result, _ := timeToExcelTime(time.Date(y, time.Month(m+1), getDaysInMonth(y, m+1), 0, 0, 0, 0, time.UTC), false)
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// HOUR function returns an integer representing the hour component of a
|
|
// supplied Excel time. The syntax of the function is:
|
|
//
|
|
// HOUR(serial_number)
|
|
func (fn *formulaFuncs) HOUR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HOUR requires exactly 1 argument")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
timeString := strings.ToLower(date.Value())
|
|
if !isTimeOnlyFmt(timeString) {
|
|
_, _, _, _, err := strToDate(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
h, _, _, pm, _, err := strToTime(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
if pm {
|
|
h += 12
|
|
}
|
|
return newNumberFormulaArg(float64(h))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "HOUR only accepts positive argument")
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Hour()))
|
|
}
|
|
|
|
// MINUTE function returns an integer representing the minute component of a
|
|
// supplied Excel time. The syntax of the function is:
|
|
//
|
|
// MINUTE(serial_number)
|
|
func (fn *formulaFuncs) MINUTE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MINUTE requires exactly 1 argument")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
timeString := strings.ToLower(date.Value())
|
|
if !isTimeOnlyFmt(timeString) {
|
|
_, _, _, _, err := strToDate(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
_, m, _, _, _, err := strToTime(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(float64(m))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "MINUTE only accepts positive argument")
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Minute()))
|
|
}
|
|
|
|
// MONTH function returns the month of a date represented by a serial number.
|
|
// The month is given as an integer, ranging from 1 (January) to 12
|
|
// (December). The syntax of the function is:
|
|
//
|
|
// MONTH(serial_number)
|
|
func (fn *formulaFuncs) MONTH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MONTH requires exactly 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(arg.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
_, month, _, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(float64(month))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "MONTH only accepts positive argument")
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Month()))
|
|
}
|
|
|
|
// genWeekendMask generate weekend mask of a series of seven 0's and 1's which
|
|
// represent the seven weekdays, starting from Monday.
|
|
func genWeekendMask(weekend int) []byte {
|
|
if masks, ok := map[int][]int{
|
|
1: {5, 6}, 2: {6, 0}, 3: {0, 1}, 4: {1, 2}, 5: {2, 3}, 6: {3, 4}, 7: {4, 5},
|
|
11: {6}, 12: {0}, 13: {1}, 14: {2}, 15: {3}, 16: {4}, 17: {5},
|
|
}[weekend]; ok {
|
|
mask := make([]byte, 7)
|
|
for _, idx := range masks {
|
|
mask[idx] = 1
|
|
}
|
|
return mask
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// isWorkday check if the date is workday.
|
|
func isWorkday(weekendMask []byte, date float64) bool {
|
|
dateTime := timeFromExcelTime(date, false)
|
|
weekday := dateTime.Weekday()
|
|
if weekday == time.Sunday {
|
|
weekday = 7
|
|
}
|
|
return weekendMask[weekday-1] == 0
|
|
}
|
|
|
|
// prepareWorkday returns weekend mask and workdays pre week by given days
|
|
// counted as weekend.
|
|
func prepareWorkday(weekend formulaArg) ([]byte, int) {
|
|
weekendArg := weekend.ToNumber()
|
|
if weekendArg.Type != ArgNumber {
|
|
return nil, 0
|
|
}
|
|
var weekendMask []byte
|
|
var workdaysPerWeek int
|
|
if len(weekend.Value()) == 7 {
|
|
// possible string values for the weekend argument
|
|
for _, mask := range weekend.Value() {
|
|
if mask != '0' && mask != '1' {
|
|
return nil, 0
|
|
}
|
|
weekendMask = append(weekendMask, byte(mask)-48)
|
|
}
|
|
} else {
|
|
weekendMask = genWeekendMask(int(weekendArg.Number))
|
|
}
|
|
for _, mask := range weekendMask {
|
|
if mask == 0 {
|
|
workdaysPerWeek++
|
|
}
|
|
}
|
|
return weekendMask, workdaysPerWeek
|
|
}
|
|
|
|
// toExcelDateArg function converts a text representation of a time, into an
|
|
// Excel date time number formula argument.
|
|
func toExcelDateArg(arg formulaArg) formulaArg {
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(arg.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
num.Number, _ = timeToExcelTime(time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.UTC), false)
|
|
return newNumberFormulaArg(num.Number)
|
|
}
|
|
if arg.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return num
|
|
}
|
|
|
|
// prepareHolidays function converts array type formula arguments to into an
|
|
// Excel date time number formula arguments list.
|
|
func prepareHolidays(args formulaArg) []int {
|
|
var holidays []int
|
|
for _, arg := range args.ToList() {
|
|
num := toExcelDateArg(arg)
|
|
if num.Type != ArgNumber {
|
|
continue
|
|
}
|
|
holidays = append(holidays, int(math.Ceil(num.Number)))
|
|
}
|
|
return holidays
|
|
}
|
|
|
|
// workdayIntl is an implementation of the formula function WORKDAY.INTL.
|
|
func workdayIntl(endDate, sign int, holidays []int, weekendMask []byte, startDate float64) int {
|
|
for i := 0; i < len(holidays); i++ {
|
|
holiday := holidays[i]
|
|
if sign > 0 {
|
|
if holiday > endDate {
|
|
break
|
|
}
|
|
} else {
|
|
if holiday < endDate {
|
|
break
|
|
}
|
|
}
|
|
if sign > 0 {
|
|
if holiday > int(math.Ceil(startDate)) {
|
|
if isWorkday(weekendMask, float64(holiday)) {
|
|
endDate += sign
|
|
for !isWorkday(weekendMask, float64(endDate)) {
|
|
endDate += sign
|
|
}
|
|
}
|
|
}
|
|
} else {
|
|
if holiday < int(math.Ceil(startDate)) {
|
|
if isWorkday(weekendMask, float64(holiday)) {
|
|
endDate += sign
|
|
for !isWorkday(weekendMask, float64(endDate)) {
|
|
endDate += sign
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return endDate
|
|
}
|
|
|
|
// NETWORKDAYS function calculates the number of work days between two supplied
|
|
// dates (including the start and end date). The calculation includes all
|
|
// weekdays (Mon - Fri), excluding a supplied list of holidays. The syntax of
|
|
// the function is:
|
|
//
|
|
// NETWORKDAYS(start_date,end_date,[holidays])
|
|
func (fn *formulaFuncs) NETWORKDAYS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NETWORKDAYS requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NETWORKDAYS requires at most 3 arguments")
|
|
}
|
|
args := list.New()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(argsList.Front().Next().Value.(formulaArg))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
if argsList.Len() == 3 {
|
|
args.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.NETWORKDAYSdotINTL(args)
|
|
}
|
|
|
|
// NETWORKDAYSdotINTL function calculates the number of whole work days between
|
|
// two supplied dates, excluding weekends and holidays. The function allows
|
|
// the user to specify which days are counted as weekends and holidays. The
|
|
// syntax of the function is:
|
|
//
|
|
// NETWORKDAYS.INTL(start_date,end_date,[weekend],[holidays])
|
|
func (fn *formulaFuncs) NETWORKDAYSdotINTL(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NETWORKDAYS.INTL requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NETWORKDAYS.INTL requires at most 4 arguments")
|
|
}
|
|
startDate := toExcelDateArg(argsList.Front().Value.(formulaArg))
|
|
if startDate.Type != ArgNumber {
|
|
return startDate
|
|
}
|
|
endDate := toExcelDateArg(argsList.Front().Next().Value.(formulaArg))
|
|
if endDate.Type != ArgNumber {
|
|
return endDate
|
|
}
|
|
weekend := newNumberFormulaArg(1)
|
|
if argsList.Len() > 2 {
|
|
weekend = argsList.Front().Next().Next().Value.(formulaArg)
|
|
}
|
|
var holidays []int
|
|
if argsList.Len() == 4 {
|
|
holidays = prepareHolidays(argsList.Back().Value.(formulaArg))
|
|
sort.Ints(holidays)
|
|
}
|
|
weekendMask, workdaysPerWeek := prepareWorkday(weekend)
|
|
if workdaysPerWeek == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
sign := 1
|
|
if startDate.Number > endDate.Number {
|
|
sign = -1
|
|
temp := startDate.Number
|
|
startDate.Number = endDate.Number
|
|
endDate.Number = temp
|
|
}
|
|
offset := endDate.Number - startDate.Number
|
|
count := int(math.Floor(offset/7) * float64(workdaysPerWeek))
|
|
daysMod := int(offset) % 7
|
|
for daysMod >= 0 {
|
|
if isWorkday(weekendMask, endDate.Number-float64(daysMod)) {
|
|
count++
|
|
}
|
|
daysMod--
|
|
}
|
|
for i := 0; i < len(holidays); i++ {
|
|
holiday := float64(holidays[i])
|
|
if isWorkday(weekendMask, holiday) && holiday >= startDate.Number && holiday <= endDate.Number {
|
|
count--
|
|
}
|
|
}
|
|
return newNumberFormulaArg(float64(sign * count))
|
|
}
|
|
|
|
// WORKDAY function returns a date that is a supplied number of working days
|
|
// (excluding weekends and holidays) ahead of a given start date. The syntax
|
|
// of the function is:
|
|
//
|
|
// WORKDAY(start_date,days,[holidays])
|
|
func (fn *formulaFuncs) WORKDAY(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WORKDAY requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WORKDAY requires at most 3 arguments")
|
|
}
|
|
args := list.New()
|
|
args.PushBack(argsList.Front().Value.(formulaArg))
|
|
args.PushBack(argsList.Front().Next().Value.(formulaArg))
|
|
args.PushBack(newNumberFormulaArg(1))
|
|
if argsList.Len() == 3 {
|
|
args.PushBack(argsList.Back().Value.(formulaArg))
|
|
}
|
|
return fn.WORKDAYdotINTL(args)
|
|
}
|
|
|
|
// WORKDAYdotINTL function returns a date that is a supplied number of working
|
|
// days (excluding weekends and holidays) ahead of a given start date. The
|
|
// function allows the user to specify which days of the week are counted as
|
|
// weekends. The syntax of the function is:
|
|
//
|
|
// WORKDAY.INTL(start_date,days,[weekend],[holidays])
|
|
func (fn *formulaFuncs) WORKDAYdotINTL(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WORKDAY.INTL requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WORKDAY.INTL requires at most 4 arguments")
|
|
}
|
|
startDate := toExcelDateArg(argsList.Front().Value.(formulaArg))
|
|
if startDate.Type != ArgNumber {
|
|
return startDate
|
|
}
|
|
days := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if days.Type != ArgNumber {
|
|
return days
|
|
}
|
|
weekend := newNumberFormulaArg(1)
|
|
if argsList.Len() > 2 {
|
|
weekend = argsList.Front().Next().Next().Value.(formulaArg)
|
|
}
|
|
var holidays []int
|
|
if argsList.Len() == 4 {
|
|
holidays = prepareHolidays(argsList.Back().Value.(formulaArg))
|
|
sort.Ints(holidays)
|
|
}
|
|
if days.Number == 0 {
|
|
return newNumberFormulaArg(math.Ceil(startDate.Number))
|
|
}
|
|
weekendMask, workdaysPerWeek := prepareWorkday(weekend)
|
|
if workdaysPerWeek == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
sign := 1
|
|
if days.Number < 0 {
|
|
sign = -1
|
|
}
|
|
offset := int(days.Number) / workdaysPerWeek
|
|
daysMod := int(days.Number) % workdaysPerWeek
|
|
endDate := int(math.Ceil(startDate.Number)) + offset*7
|
|
if daysMod == 0 {
|
|
for !isWorkday(weekendMask, float64(endDate)) {
|
|
endDate -= sign
|
|
}
|
|
} else {
|
|
for daysMod != 0 {
|
|
endDate += sign
|
|
if isWorkday(weekendMask, float64(endDate)) {
|
|
if daysMod < 0 {
|
|
daysMod++
|
|
continue
|
|
}
|
|
daysMod--
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(float64(workdayIntl(endDate, sign, holidays, weekendMask, startDate.Number)))
|
|
}
|
|
|
|
// YEAR function returns an integer representing the year of a supplied date.
|
|
// The syntax of the function is:
|
|
//
|
|
// YEAR(serial_number)
|
|
func (fn *formulaFuncs) YEAR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "YEAR requires exactly 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg)
|
|
num := arg.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(arg.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
year, _, _, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(float64(year))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YEAR only accepts positive argument")
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Year()))
|
|
}
|
|
|
|
// yearFracBasisCond is an implementation of the yearFracBasis1.
|
|
func yearFracBasisCond(sy, sm, sd, ey, em, ed int) bool {
|
|
return (isLeapYear(sy) && (sm < 2 || (sm == 2 && sd <= 29))) || (isLeapYear(ey) && (em > 2 || (em == 2 && ed == 29)))
|
|
}
|
|
|
|
// yearFracBasis0 function returns the fraction of a year that between two
|
|
// supplied dates in US (NASD) 30/360 type of day.
|
|
func yearFracBasis0(startDate, endDate float64) (dayDiff, daysInYear float64) {
|
|
startTime, endTime := timeFromExcelTime(startDate, false), timeFromExcelTime(endDate, false)
|
|
sy, smM, sd := startTime.Date()
|
|
ey, emM, ed := endTime.Date()
|
|
sm, em := int(smM), int(emM)
|
|
if sd == 31 {
|
|
sd--
|
|
}
|
|
if sd == 30 && ed == 31 {
|
|
ed--
|
|
} else if leap := isLeapYear(sy); sm == 2 && ((leap && sd == 29) || (!leap && sd == 28)) {
|
|
sd = 30
|
|
if leap := isLeapYear(ey); em == 2 && ((leap && ed == 29) || (!leap && ed == 28)) {
|
|
ed = 30
|
|
}
|
|
}
|
|
dayDiff = float64((ey-sy)*360 + (em-sm)*30 + (ed - sd))
|
|
daysInYear = 360
|
|
return
|
|
}
|
|
|
|
// yearFracBasis1 function returns the fraction of a year that between two
|
|
// supplied dates in actual type of day.
|
|
func yearFracBasis1(startDate, endDate float64) (dayDiff, daysInYear float64) {
|
|
startTime, endTime := timeFromExcelTime(startDate, false), timeFromExcelTime(endDate, false)
|
|
sy, smM, sd := startTime.Date()
|
|
ey, emM, ed := endTime.Date()
|
|
sm, em := int(smM), int(emM)
|
|
dayDiff = endDate - startDate
|
|
isYearDifferent := sy != ey
|
|
if isYearDifferent && (ey != sy+1 || sm < em || (sm == em && sd < ed)) {
|
|
dayCount := 0
|
|
for y := sy; y <= ey; y++ {
|
|
dayCount += getYearDays(y, 1)
|
|
}
|
|
daysInYear = float64(dayCount) / float64(ey-sy+1)
|
|
} else {
|
|
if !isYearDifferent && isLeapYear(sy) {
|
|
daysInYear = 366
|
|
} else {
|
|
if isYearDifferent && yearFracBasisCond(sy, sm, sd, ey, em, ed) {
|
|
daysInYear = 366
|
|
} else {
|
|
daysInYear = 365
|
|
}
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// yearFracBasis4 function returns the fraction of a year that between two
|
|
// supplied dates in European 30/360 type of day.
|
|
func yearFracBasis4(startDate, endDate float64) (dayDiff, daysInYear float64) {
|
|
startTime, endTime := timeFromExcelTime(startDate, false), timeFromExcelTime(endDate, false)
|
|
sy, smM, sd := startTime.Date()
|
|
ey, emM, ed := endTime.Date()
|
|
sm, em := int(smM), int(emM)
|
|
if sd == 31 {
|
|
sd--
|
|
}
|
|
if ed == 31 {
|
|
ed--
|
|
}
|
|
dayDiff = float64((ey-sy)*360 + (em-sm)*30 + (ed - sd))
|
|
daysInYear = 360
|
|
return
|
|
}
|
|
|
|
// yearFrac is an implementation of the formula function YEARFRAC.
|
|
func yearFrac(startDate, endDate float64, basis int) formulaArg {
|
|
startTime, endTime := timeFromExcelTime(startDate, false), timeFromExcelTime(endDate, false)
|
|
if startTime == endTime {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
var dayDiff, daysInYear float64
|
|
switch basis {
|
|
case 0:
|
|
dayDiff, daysInYear = yearFracBasis0(startDate, endDate)
|
|
case 1:
|
|
dayDiff, daysInYear = yearFracBasis1(startDate, endDate)
|
|
case 2:
|
|
dayDiff = endDate - startDate
|
|
daysInYear = 360
|
|
case 3:
|
|
dayDiff = endDate - startDate
|
|
daysInYear = 365
|
|
case 4:
|
|
dayDiff, daysInYear = yearFracBasis4(startDate, endDate)
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorNUM, "invalid basis")
|
|
}
|
|
return newNumberFormulaArg(dayDiff / daysInYear)
|
|
}
|
|
|
|
// getYearDays return days of the year with specifying the type of day count
|
|
// basis to be used.
|
|
func getYearDays(year, basis int) int {
|
|
switch basis {
|
|
case 1:
|
|
if isLeapYear(year) {
|
|
return 366
|
|
}
|
|
return 365
|
|
case 3:
|
|
return 365
|
|
default:
|
|
return 360
|
|
}
|
|
}
|
|
|
|
// YEARFRAC function returns the fraction of a year that is represented by the
|
|
// number of whole days between two supplied dates. The syntax of the
|
|
// function is:
|
|
//
|
|
// YEARFRAC(start_date,end_date,[basis])
|
|
func (fn *formulaFuncs) YEARFRAC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 && argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "YEARFRAC requires 3 or 4 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
start, end := args.List[0], args.List[1]
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 3 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return basis
|
|
}
|
|
}
|
|
return yearFrac(start.Number, end.Number, int(basis.Number))
|
|
}
|
|
|
|
// NOW function returns the current date and time. The function receives no
|
|
// arguments and therefore. The syntax of the function is:
|
|
//
|
|
// NOW()
|
|
func (fn *formulaFuncs) NOW(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NOW accepts no arguments")
|
|
}
|
|
now := time.Now()
|
|
_, offset := now.Zone()
|
|
return newNumberFormulaArg(25569.0 + float64(now.Unix()+int64(offset))/86400)
|
|
}
|
|
|
|
// SECOND function returns an integer representing the second component of a
|
|
// supplied Excel time. The syntax of the function is:
|
|
//
|
|
// SECOND(serial_number)
|
|
func (fn *formulaFuncs) SECOND(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SECOND requires exactly 1 argument")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
num := date.ToNumber()
|
|
if num.Type != ArgNumber {
|
|
timeString := strings.ToLower(date.Value())
|
|
if !isTimeOnlyFmt(timeString) {
|
|
_, _, _, _, err := strToDate(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
_, _, s, _, _, err := strToTime(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
return newNumberFormulaArg(float64(int(s) % 60))
|
|
}
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "SECOND only accepts positive argument")
|
|
}
|
|
return newNumberFormulaArg(float64(timeFromExcelTime(num.Number, false).Second()))
|
|
}
|
|
|
|
// TIME function accepts three integer arguments representing hours, minutes
|
|
// and seconds, and returns an Excel time. I.e. the function returns the
|
|
// decimal value that represents the time in Excel. The syntax of the
|
|
// function is:
|
|
//
|
|
// TIME(hour,minute,second)
|
|
func (fn *formulaFuncs) TIME(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TIME requires 3 number arguments")
|
|
}
|
|
h := argsList.Front().Value.(formulaArg).ToNumber()
|
|
m := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
s := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if h.Type != ArgNumber || m.Type != ArgNumber || s.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TIME requires 3 number arguments")
|
|
}
|
|
t := (h.Number*3600 + m.Number*60 + s.Number) / 86400
|
|
if t < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(t)
|
|
}
|
|
|
|
// TIMEVALUE function converts a text representation of a time, into an Excel
|
|
// time. The syntax of the function is:
|
|
//
|
|
// TIMEVALUE(time_text)
|
|
func (fn *formulaFuncs) TIMEVALUE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TIMEVALUE requires exactly 1 argument")
|
|
}
|
|
date := argsList.Front().Value.(formulaArg)
|
|
timeString := strings.ToLower(date.Value())
|
|
if !isTimeOnlyFmt(timeString) {
|
|
_, _, _, _, err := strToDate(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
h, m, s, pm, _, err := strToTime(timeString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
if pm {
|
|
h += 12
|
|
}
|
|
args := list.New()
|
|
args.PushBack(newNumberFormulaArg(float64(h)))
|
|
args.PushBack(newNumberFormulaArg(float64(m)))
|
|
args.PushBack(newNumberFormulaArg(s))
|
|
return fn.TIME(args)
|
|
}
|
|
|
|
// TODAY function returns the current date. The function has no arguments and
|
|
// therefore. The syntax of the function is:
|
|
//
|
|
// TODAY()
|
|
func (fn *formulaFuncs) TODAY(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TODAY accepts no arguments")
|
|
}
|
|
now := time.Now()
|
|
_, offset := now.Zone()
|
|
return newNumberFormulaArg(daysBetween(excelMinTime1900.Unix(), now.Unix()+int64(offset)) + 1)
|
|
}
|
|
|
|
// makeDate return date as a Unix time, the number of seconds elapsed since
|
|
// January 1, 1970 UTC.
|
|
func makeDate(y int, m time.Month, d int) int64 {
|
|
if y == 1900 && int(m) <= 2 {
|
|
d--
|
|
}
|
|
date := time.Date(y, m, d, 0, 0, 0, 0, time.UTC)
|
|
return date.Unix()
|
|
}
|
|
|
|
// daysBetween return time interval of the given start timestamp and end
|
|
// timestamp.
|
|
func daysBetween(startDate, endDate int64) float64 {
|
|
return float64(int(0.5 + float64((endDate-startDate)/86400)))
|
|
}
|
|
|
|
// WEEKDAY function returns an integer representing the day of the week for a
|
|
// supplied date. The syntax of the function is:
|
|
//
|
|
// WEEKDAY(serial_number,[return_type])
|
|
func (fn *formulaFuncs) WEEKDAY(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEEKDAY requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEEKDAY allows at most 2 arguments")
|
|
}
|
|
sn := argsList.Front().Value.(formulaArg)
|
|
num := sn.ToNumber()
|
|
weekday, returnType := 0, 1
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(sn.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
weekday = int(time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.Now().Location()).Weekday())
|
|
} else {
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
weekday = int(timeFromExcelTime(num.Number, false).Weekday())
|
|
}
|
|
if argsList.Len() == 2 {
|
|
returnTypeArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if returnTypeArg.Type != ArgNumber {
|
|
return returnTypeArg
|
|
}
|
|
returnType = int(returnTypeArg.Number)
|
|
}
|
|
if returnType == 2 {
|
|
returnType = 11
|
|
}
|
|
weekday++
|
|
if returnType == 1 {
|
|
return newNumberFormulaArg(float64(weekday))
|
|
}
|
|
if returnType == 3 {
|
|
return newNumberFormulaArg(float64((weekday + 6 - 1) % 7))
|
|
}
|
|
if returnType >= 11 && returnType <= 17 {
|
|
return newNumberFormulaArg(float64((weekday+6-(returnType-10))%7 + 1))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
|
|
// weeknum is an implementation of the formula function WEEKNUM.
|
|
func (fn *formulaFuncs) weeknum(snTime time.Time, returnType int) formulaArg {
|
|
days := snTime.YearDay()
|
|
weekMod, weekNum := days%7, math.Ceil(float64(days)/7)
|
|
if weekMod == 0 {
|
|
weekMod = 7
|
|
}
|
|
year := snTime.Year()
|
|
firstWeekday := int(time.Date(year, time.January, 1, 0, 0, 0, 0, time.UTC).Weekday())
|
|
var offset int
|
|
switch returnType {
|
|
case 1, 17:
|
|
offset = 0
|
|
case 2, 11, 21:
|
|
offset = 1
|
|
case 12, 13, 14, 15, 16:
|
|
offset = returnType - 10
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
padding := offset + 7 - firstWeekday
|
|
if padding > 7 {
|
|
padding -= 7
|
|
}
|
|
if weekMod > padding {
|
|
weekNum++
|
|
}
|
|
if returnType == 21 && (firstWeekday == 0 || firstWeekday > 4) {
|
|
if weekNum--; weekNum < 1 {
|
|
if weekNum = 52; int(time.Date(year-1, time.January, 1, 0, 0, 0, 0, time.UTC).Weekday()) < 4 {
|
|
weekNum++
|
|
}
|
|
}
|
|
}
|
|
return newNumberFormulaArg(weekNum)
|
|
}
|
|
|
|
// WEEKNUM function returns an integer representing the week number (from 1 to
|
|
// 53) of the year. The syntax of the function is:
|
|
//
|
|
// WEEKNUM(serial_number,[return_type])
|
|
func (fn *formulaFuncs) WEEKNUM(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEEKNUM requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "WEEKNUM allows at most 2 arguments")
|
|
}
|
|
sn := argsList.Front().Value.(formulaArg)
|
|
num, returnType := sn.ToNumber(), 1
|
|
var snTime time.Time
|
|
if num.Type != ArgNumber {
|
|
dateString := strings.ToLower(sn.Value())
|
|
if !isDateOnlyFmt(dateString) {
|
|
if _, _, _, _, _, err := strToTime(dateString); err.Type == ArgError {
|
|
return err
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(dateString)
|
|
if err.Type == ArgError {
|
|
return err
|
|
}
|
|
snTime = time.Date(y, time.Month(m), d, 0, 0, 0, 0, time.Now().Location())
|
|
} else {
|
|
if num.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
snTime = timeFromExcelTime(num.Number, false)
|
|
}
|
|
if argsList.Len() == 2 {
|
|
returnTypeArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if returnTypeArg.Type != ArgNumber {
|
|
return returnTypeArg
|
|
}
|
|
returnType = int(returnTypeArg.Number)
|
|
}
|
|
return fn.weeknum(snTime, returnType)
|
|
}
|
|
|
|
// Text Functions
|
|
|
|
// CHAR function returns the character relating to a supplied character set
|
|
// number (from 1 to 255). syntax of the function is:
|
|
//
|
|
// CHAR(number)
|
|
func (fn *formulaFuncs) CHAR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHAR requires 1 argument")
|
|
}
|
|
arg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if arg.Type != ArgNumber {
|
|
return arg
|
|
}
|
|
num := int(arg.Number)
|
|
if num < 0 || num > MaxFieldLength {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newStringFormulaArg(fmt.Sprintf("%c", num))
|
|
}
|
|
|
|
// CLEAN removes all non-printable characters from a supplied text string. The
|
|
// syntax of the function is:
|
|
//
|
|
// CLEAN(text)
|
|
func (fn *formulaFuncs) CLEAN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CLEAN requires 1 argument")
|
|
}
|
|
b := bytes.Buffer{}
|
|
for _, c := range argsList.Front().Value.(formulaArg).Value() {
|
|
if c > 31 {
|
|
b.WriteRune(c)
|
|
}
|
|
}
|
|
return newStringFormulaArg(b.String())
|
|
}
|
|
|
|
// CODE function converts the first character of a supplied text string into
|
|
// the associated numeric character set code used by your computer. The
|
|
// syntax of the function is:
|
|
//
|
|
// CODE(text)
|
|
func (fn *formulaFuncs) CODE(argsList *list.List) formulaArg {
|
|
return fn.code("CODE", argsList)
|
|
}
|
|
|
|
// code is an implementation of the formula functions CODE and UNICODE.
|
|
func (fn *formulaFuncs) code(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 1 argument", name))
|
|
}
|
|
text := argsList.Front().Value.(formulaArg).Value()
|
|
if len(text) == 0 {
|
|
if name == "CODE" {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newNumberFormulaArg(float64(text[0]))
|
|
}
|
|
|
|
// CONCAT function joins together a series of supplied text strings into one
|
|
// combined text string.
|
|
//
|
|
// CONCAT(text1,[text2],...)
|
|
func (fn *formulaFuncs) CONCAT(argsList *list.List) formulaArg {
|
|
return fn.concat("CONCAT", argsList)
|
|
}
|
|
|
|
// CONCATENATE function joins together a series of supplied text strings into
|
|
// one combined text string.
|
|
//
|
|
// CONCATENATE(text1,[text2],...)
|
|
func (fn *formulaFuncs) CONCATENATE(argsList *list.List) formulaArg {
|
|
return fn.concat("CONCATENATE", argsList)
|
|
}
|
|
|
|
// concat is an implementation of the formula functions CONCAT and
|
|
// CONCATENATE.
|
|
func (fn *formulaFuncs) concat(name string, argsList *list.List) formulaArg {
|
|
buf := bytes.Buffer{}
|
|
for arg := argsList.Front(); arg != nil; arg = arg.Next() {
|
|
token := arg.Value.(formulaArg)
|
|
switch token.Type {
|
|
case ArgString:
|
|
buf.WriteString(token.String)
|
|
case ArgNumber:
|
|
if token.Boolean {
|
|
if token.Number == 0 {
|
|
buf.WriteString("FALSE")
|
|
} else {
|
|
buf.WriteString("TRUE")
|
|
}
|
|
} else {
|
|
buf.WriteString(token.Value())
|
|
}
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires arguments to be strings", name))
|
|
}
|
|
}
|
|
return newStringFormulaArg(buf.String())
|
|
}
|
|
|
|
// EXACT function tests if two supplied text strings or values are exactly
|
|
// equal and if so, returns TRUE; Otherwise, the function returns FALSE. The
|
|
// function is case-sensitive. The syntax of the function is:
|
|
//
|
|
// EXACT(text1,text2)
|
|
func (fn *formulaFuncs) EXACT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EXACT requires 2 arguments")
|
|
}
|
|
text1 := argsList.Front().Value.(formulaArg).Value()
|
|
text2 := argsList.Back().Value.(formulaArg).Value()
|
|
return newBoolFormulaArg(text1 == text2)
|
|
}
|
|
|
|
// FIXED function rounds a supplied number to a specified number of decimal
|
|
// places and then converts this into text. The syntax of the function is:
|
|
//
|
|
// FIXED(number,[decimals],[no_commas])
|
|
func (fn *formulaFuncs) FIXED(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FIXED requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FIXED allows at most 3 arguments")
|
|
}
|
|
numArg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if numArg.Type != ArgNumber {
|
|
return numArg
|
|
}
|
|
precision, decimals, noCommas := 0, 0, false
|
|
s := strings.Split(argsList.Front().Value.(formulaArg).Value(), ".")
|
|
if argsList.Len() == 1 && len(s) == 2 {
|
|
precision = len(s[1])
|
|
decimals = len(s[1])
|
|
}
|
|
if argsList.Len() >= 2 {
|
|
decimalsArg := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if decimalsArg.Type != ArgNumber {
|
|
return decimalsArg
|
|
}
|
|
decimals = int(decimalsArg.Number)
|
|
}
|
|
if argsList.Len() == 3 {
|
|
noCommasArg := argsList.Back().Value.(formulaArg).ToBool()
|
|
if noCommasArg.Type == ArgError {
|
|
return noCommasArg
|
|
}
|
|
noCommas = noCommasArg.Boolean
|
|
}
|
|
n := math.Pow(10, float64(decimals))
|
|
r := numArg.Number * n
|
|
fixed := float64(int(r+math.Copysign(0.5, r))) / n
|
|
if decimals > 0 {
|
|
precision = decimals
|
|
}
|
|
if noCommas {
|
|
return newStringFormulaArg(fmt.Sprintf(fmt.Sprintf("%%.%df", precision), fixed))
|
|
}
|
|
p := message.NewPrinter(language.English)
|
|
return newStringFormulaArg(p.Sprintf(fmt.Sprintf("%%.%df", precision), fixed))
|
|
}
|
|
|
|
// FIND function returns the position of a specified character or sub-string
|
|
// within a supplied text string. The function is case-sensitive. The syntax
|
|
// of the function is:
|
|
//
|
|
// FIND(find_text,within_text,[start_num])
|
|
func (fn *formulaFuncs) FIND(argsList *list.List) formulaArg {
|
|
return fn.find("FIND", argsList)
|
|
}
|
|
|
|
// FINDB counts each double-byte character as 2 when you have enabled the
|
|
// editing of a language that supports DBCS and then set it as the default
|
|
// language. Otherwise, FINDB counts each character as 1. The syntax of the
|
|
// function is:
|
|
//
|
|
// FINDB(find_text,within_text,[start_num])
|
|
func (fn *formulaFuncs) FINDB(argsList *list.List) formulaArg {
|
|
return fn.find("FINDB", argsList)
|
|
}
|
|
|
|
// find is an implementation of the formula functions FIND and FINDB.
|
|
func (fn *formulaFuncs) find(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 2 arguments", name))
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 3 arguments", name))
|
|
}
|
|
findText := argsList.Front().Value.(formulaArg).Value()
|
|
withinText := argsList.Front().Next().Value.(formulaArg).Value()
|
|
startNum, result := 1, 1
|
|
if argsList.Len() == 3 {
|
|
numArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if numArg.Type != ArgNumber {
|
|
return numArg
|
|
}
|
|
if numArg.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
startNum = int(numArg.Number)
|
|
}
|
|
if findText == "" {
|
|
return newNumberFormulaArg(float64(startNum))
|
|
}
|
|
for idx := range withinText {
|
|
if result < startNum {
|
|
result++
|
|
}
|
|
if strings.Index(withinText[idx:], findText) == 0 {
|
|
return newNumberFormulaArg(float64(result))
|
|
}
|
|
result++
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
|
|
// LEFT function returns a specified number of characters from the start of a
|
|
// supplied text string. The syntax of the function is:
|
|
//
|
|
// LEFT(text,[num_chars])
|
|
func (fn *formulaFuncs) LEFT(argsList *list.List) formulaArg {
|
|
return fn.leftRight("LEFT", argsList)
|
|
}
|
|
|
|
// LEFTB returns the first character or characters in a text string, based on
|
|
// the number of bytes you specify. The syntax of the function is:
|
|
//
|
|
// LEFTB(text,[num_bytes])
|
|
func (fn *formulaFuncs) LEFTB(argsList *list.List) formulaArg {
|
|
return fn.leftRight("LEFTB", argsList)
|
|
}
|
|
|
|
// leftRight is an implementation of the formula functions LEFT, LEFTB, RIGHT,
|
|
// RIGHTB. TODO: support DBCS include Japanese, Chinese (Simplified), Chinese
|
|
// (Traditional), and Korean.
|
|
func (fn *formulaFuncs) leftRight(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 1 argument", name))
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 2 arguments", name))
|
|
}
|
|
text, numChars := argsList.Front().Value.(formulaArg).Value(), 1
|
|
if argsList.Len() == 2 {
|
|
numArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if numArg.Type != ArgNumber {
|
|
return numArg
|
|
}
|
|
if numArg.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
numChars = int(numArg.Number)
|
|
}
|
|
if utf8.RuneCountInString(text) > numChars {
|
|
if name == "LEFT" || name == "LEFTB" {
|
|
return newStringFormulaArg(string([]rune(text)[:numChars]))
|
|
}
|
|
return newStringFormulaArg(string([]rune(text)[utf8.RuneCountInString(text)-numChars:]))
|
|
}
|
|
return newStringFormulaArg(text)
|
|
}
|
|
|
|
// LEN returns the length of a supplied text string. The syntax of the
|
|
// function is:
|
|
//
|
|
// LEN(text)
|
|
func (fn *formulaFuncs) LEN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LEN requires 1 string argument")
|
|
}
|
|
return newStringFormulaArg(strconv.Itoa(utf8.RuneCountInString(argsList.Front().Value.(formulaArg).String)))
|
|
}
|
|
|
|
// LENB returns the number of bytes used to represent the characters in a text
|
|
// string. LENB counts 2 bytes per character only when a DBCS language is set
|
|
// as the default language. Otherwise LENB behaves the same as LEN, counting
|
|
// 1 byte per character. The syntax of the function is:
|
|
//
|
|
// LENB(text)
|
|
//
|
|
// TODO: the languages that support DBCS include Japanese, Chinese
|
|
// (Simplified), Chinese (Traditional), and Korean.
|
|
func (fn *formulaFuncs) LENB(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LENB requires 1 string argument")
|
|
}
|
|
return newStringFormulaArg(strconv.Itoa(len(argsList.Front().Value.(formulaArg).String)))
|
|
}
|
|
|
|
// LOWER converts all characters in a supplied text string to lower case. The
|
|
// syntax of the function is:
|
|
//
|
|
// LOWER(text)
|
|
func (fn *formulaFuncs) LOWER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "LOWER requires 1 argument")
|
|
}
|
|
return newStringFormulaArg(strings.ToLower(argsList.Front().Value.(formulaArg).String))
|
|
}
|
|
|
|
// MID function returns a specified number of characters from the middle of a
|
|
// supplied text string. The syntax of the function is:
|
|
//
|
|
// MID(text,start_num,num_chars)
|
|
func (fn *formulaFuncs) MID(argsList *list.List) formulaArg {
|
|
return fn.mid("MID", argsList)
|
|
}
|
|
|
|
// MIDB returns a specific number of characters from a text string, starting
|
|
// at the position you specify, based on the number of bytes you specify. The
|
|
// syntax of the function is:
|
|
//
|
|
// MID(text,start_num,num_chars)
|
|
func (fn *formulaFuncs) MIDB(argsList *list.List) formulaArg {
|
|
return fn.mid("MIDB", argsList)
|
|
}
|
|
|
|
// mid is an implementation of the formula functions MID and MIDB.
|
|
func (fn *formulaFuncs) mid(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 3 arguments", name))
|
|
}
|
|
text := argsList.Front().Value.(formulaArg).Value()
|
|
startNumArg, numCharsArg := argsList.Front().Next().Value.(formulaArg).ToNumber(), argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if startNumArg.Type != ArgNumber {
|
|
return startNumArg
|
|
}
|
|
if numCharsArg.Type != ArgNumber {
|
|
return numCharsArg
|
|
}
|
|
startNum := int(startNumArg.Number)
|
|
if startNum < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
textLen := utf8.RuneCountInString(text)
|
|
if startNum > textLen {
|
|
return newStringFormulaArg("")
|
|
}
|
|
startNum--
|
|
endNum := startNum + int(numCharsArg.Number)
|
|
if endNum > textLen+1 {
|
|
return newStringFormulaArg(text[startNum:])
|
|
}
|
|
return newStringFormulaArg(text[startNum:endNum])
|
|
}
|
|
|
|
// PROPER converts all characters in a supplied text string to proper case
|
|
// (i.e. all letters that do not immediately follow another letter are set to
|
|
// upper case and all other characters are lower case). The syntax of the
|
|
// function is:
|
|
//
|
|
// PROPER(text)
|
|
func (fn *formulaFuncs) PROPER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PROPER requires 1 argument")
|
|
}
|
|
buf := bytes.Buffer{}
|
|
isLetter := false
|
|
for _, char := range argsList.Front().Value.(formulaArg).String {
|
|
if !isLetter && unicode.IsLetter(char) {
|
|
buf.WriteRune(unicode.ToUpper(char))
|
|
} else {
|
|
buf.WriteRune(unicode.ToLower(char))
|
|
}
|
|
isLetter = unicode.IsLetter(char)
|
|
}
|
|
return newStringFormulaArg(buf.String())
|
|
}
|
|
|
|
// REPLACE function replaces all or part of a text string with another string.
|
|
// The syntax of the function is:
|
|
//
|
|
// REPLACE(old_text,start_num,num_chars,new_text)
|
|
func (fn *formulaFuncs) REPLACE(argsList *list.List) formulaArg {
|
|
return fn.replace("REPLACE", argsList)
|
|
}
|
|
|
|
// REPLACEB replaces part of a text string, based on the number of bytes you
|
|
// specify, with a different text string.
|
|
//
|
|
// REPLACEB(old_text,start_num,num_chars,new_text)
|
|
func (fn *formulaFuncs) REPLACEB(argsList *list.List) formulaArg {
|
|
return fn.replace("REPLACEB", argsList)
|
|
}
|
|
|
|
// replace is an implementation of the formula functions REPLACE and REPLACEB.
|
|
// TODO: support DBCS include Japanese, Chinese (Simplified), Chinese
|
|
// (Traditional), and Korean.
|
|
func (fn *formulaFuncs) replace(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 4 arguments", name))
|
|
}
|
|
sourceText, targetText := argsList.Front().Value.(formulaArg).Value(), argsList.Back().Value.(formulaArg).Value()
|
|
startNumArg, numCharsArg := argsList.Front().Next().Value.(formulaArg).ToNumber(), argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if startNumArg.Type != ArgNumber {
|
|
return startNumArg
|
|
}
|
|
if numCharsArg.Type != ArgNumber {
|
|
return numCharsArg
|
|
}
|
|
sourceTextLen, startIdx := len(sourceText), int(startNumArg.Number)
|
|
if startIdx > sourceTextLen {
|
|
startIdx = sourceTextLen + 1
|
|
}
|
|
endIdx := startIdx + int(numCharsArg.Number)
|
|
if endIdx > sourceTextLen {
|
|
endIdx = sourceTextLen + 1
|
|
}
|
|
if startIdx < 1 || endIdx < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
result := sourceText[:startIdx-1] + targetText + sourceText[endIdx-1:]
|
|
return newStringFormulaArg(result)
|
|
}
|
|
|
|
// REPT function returns a supplied text string, repeated a specified number
|
|
// of times. The syntax of the function is:
|
|
//
|
|
// REPT(text,number_times)
|
|
func (fn *formulaFuncs) REPT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "REPT requires 2 arguments")
|
|
}
|
|
text := argsList.Front().Value.(formulaArg)
|
|
if text.Type != ArgString {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "REPT requires first argument to be a string")
|
|
}
|
|
times := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if times.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "REPT requires second argument to be a number")
|
|
}
|
|
if times.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "REPT requires second argument to be >= 0")
|
|
}
|
|
if times.Number == 0 {
|
|
return newStringFormulaArg("")
|
|
}
|
|
buf := bytes.Buffer{}
|
|
for i := 0; i < int(times.Number); i++ {
|
|
buf.WriteString(text.String)
|
|
}
|
|
return newStringFormulaArg(buf.String())
|
|
}
|
|
|
|
// RIGHT function returns a specified number of characters from the end of a
|
|
// supplied text string. The syntax of the function is:
|
|
//
|
|
// RIGHT(text,[num_chars])
|
|
func (fn *formulaFuncs) RIGHT(argsList *list.List) formulaArg {
|
|
return fn.leftRight("RIGHT", argsList)
|
|
}
|
|
|
|
// RIGHTB returns the last character or characters in a text string, based on
|
|
// the number of bytes you specify. The syntax of the function is:
|
|
//
|
|
// RIGHTB(text,[num_bytes])
|
|
func (fn *formulaFuncs) RIGHTB(argsList *list.List) formulaArg {
|
|
return fn.leftRight("RIGHTB", argsList)
|
|
}
|
|
|
|
// SUBSTITUTE function replaces one or more instances of a given text string,
|
|
// within an original text string. The syntax of the function is:
|
|
//
|
|
// SUBSTITUTE(text,old_text,new_text,[instance_num])
|
|
func (fn *formulaFuncs) SUBSTITUTE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 && argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SUBSTITUTE requires 3 or 4 arguments")
|
|
}
|
|
text, sourceText := argsList.Front().Value.(formulaArg), argsList.Front().Next().Value.(formulaArg)
|
|
targetText, instanceNum := argsList.Front().Next().Next().Value.(formulaArg), 0
|
|
if argsList.Len() == 3 {
|
|
return newStringFormulaArg(strings.ReplaceAll(text.Value(), sourceText.Value(), targetText.Value()))
|
|
}
|
|
instanceNumArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if instanceNumArg.Type != ArgNumber {
|
|
return instanceNumArg
|
|
}
|
|
instanceNum = int(instanceNumArg.Number)
|
|
if instanceNum < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "instance_num should be > 0")
|
|
}
|
|
str, sourceTextLen, count, chars, pos := text.Value(), len(sourceText.Value()), instanceNum, 0, -1
|
|
for {
|
|
count--
|
|
index := strings.Index(str, sourceText.Value())
|
|
if index == -1 {
|
|
pos = -1
|
|
break
|
|
} else {
|
|
pos = index + chars
|
|
if count == 0 {
|
|
break
|
|
}
|
|
idx := sourceTextLen + index
|
|
chars += idx
|
|
str = str[idx:]
|
|
}
|
|
}
|
|
if pos == -1 {
|
|
return newStringFormulaArg(text.Value())
|
|
}
|
|
pre, post := text.Value()[:pos], text.Value()[pos+sourceTextLen:]
|
|
return newStringFormulaArg(pre + targetText.Value() + post)
|
|
}
|
|
|
|
// TEXTJOIN function joins together a series of supplied text strings into one
|
|
// combined text string. The user can specify a delimiter to add between the
|
|
// individual text items, if required. The syntax of the function is:
|
|
//
|
|
// TEXTJOIN([delimiter],[ignore_empty],text1,[text2],...)
|
|
func (fn *formulaFuncs) TEXTJOIN(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TEXTJOIN requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 252 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TEXTJOIN accepts at most 252 arguments")
|
|
}
|
|
delimiter := argsList.Front().Value.(formulaArg)
|
|
ignoreEmpty := argsList.Front().Next().Value.(formulaArg)
|
|
if ignoreEmpty.Type != ArgNumber || !ignoreEmpty.Boolean {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
args, ok := textJoin(argsList.Front().Next().Next(), []string{}, ignoreEmpty.Number != 0)
|
|
if ok.Type != ArgNumber {
|
|
return ok
|
|
}
|
|
result := strings.Join(args, delimiter.Value())
|
|
if len(result) > TotalCellChars {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("TEXTJOIN function exceeds %d characters", TotalCellChars))
|
|
}
|
|
return newStringFormulaArg(result)
|
|
}
|
|
|
|
// textJoin is an implementation of the formula function TEXTJOIN.
|
|
func textJoin(arg *list.Element, arr []string, ignoreEmpty bool) ([]string, formulaArg) {
|
|
for arg.Next(); arg != nil; arg = arg.Next() {
|
|
switch arg.Value.(formulaArg).Type {
|
|
case ArgError:
|
|
return arr, arg.Value.(formulaArg)
|
|
case ArgString, ArgEmpty:
|
|
val := arg.Value.(formulaArg).Value()
|
|
if val != "" || !ignoreEmpty {
|
|
arr = append(arr, val)
|
|
}
|
|
case ArgNumber:
|
|
arr = append(arr, arg.Value.(formulaArg).Value())
|
|
case ArgMatrix:
|
|
for _, row := range arg.Value.(formulaArg).Matrix {
|
|
argList := list.New().Init()
|
|
for _, ele := range row {
|
|
argList.PushBack(ele)
|
|
}
|
|
if argList.Len() > 0 {
|
|
args, _ := textJoin(argList.Front(), []string{}, ignoreEmpty)
|
|
arr = append(arr, args...)
|
|
}
|
|
}
|
|
}
|
|
}
|
|
return arr, newBoolFormulaArg(true)
|
|
}
|
|
|
|
// TRIM removes extra spaces (i.e. all spaces except for single spaces between
|
|
// words or characters) from a supplied text string. The syntax of the
|
|
// function is:
|
|
//
|
|
// TRIM(text)
|
|
func (fn *formulaFuncs) TRIM(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TRIM requires 1 argument")
|
|
}
|
|
return newStringFormulaArg(strings.TrimSpace(argsList.Front().Value.(formulaArg).Value()))
|
|
}
|
|
|
|
// UNICHAR returns the Unicode character that is referenced by the given
|
|
// numeric value. The syntax of the function is:
|
|
//
|
|
// UNICHAR(number)
|
|
func (fn *formulaFuncs) UNICHAR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "UNICHAR requires 1 argument")
|
|
}
|
|
numArg := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if numArg.Type != ArgNumber {
|
|
return numArg
|
|
}
|
|
if numArg.Number <= 0 || numArg.Number > 55295 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newStringFormulaArg(string(rune(numArg.Number)))
|
|
}
|
|
|
|
// UNICODE function returns the code point for the first character of a
|
|
// supplied text string. The syntax of the function is:
|
|
//
|
|
// UNICODE(text)
|
|
func (fn *formulaFuncs) UNICODE(argsList *list.List) formulaArg {
|
|
return fn.code("UNICODE", argsList)
|
|
}
|
|
|
|
// UPPER converts all characters in a supplied text string to upper case. The
|
|
// syntax of the function is:
|
|
//
|
|
// UPPER(text)
|
|
func (fn *formulaFuncs) UPPER(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "UPPER requires 1 argument")
|
|
}
|
|
return newStringFormulaArg(strings.ToUpper(argsList.Front().Value.(formulaArg).String))
|
|
}
|
|
|
|
// VALUE function converts a text string into a numeric value. The syntax of
|
|
// the function is:
|
|
//
|
|
// VALUE(text)
|
|
func (fn *formulaFuncs) VALUE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "VALUE requires 1 argument")
|
|
}
|
|
text := strings.ReplaceAll(argsList.Front().Value.(formulaArg).Value(), ",", "")
|
|
percent := 1.0
|
|
if strings.HasSuffix(text, "%") {
|
|
percent, text = 0.01, strings.TrimSuffix(text, "%")
|
|
}
|
|
decimal := big.Float{}
|
|
if _, ok := decimal.SetString(text); ok {
|
|
value, _ := decimal.Float64()
|
|
return newNumberFormulaArg(value * percent)
|
|
}
|
|
dateValue, timeValue, errTime, errDate := 0.0, 0.0, false, false
|
|
if !isDateOnlyFmt(text) {
|
|
h, m, s, _, _, err := strToTime(text)
|
|
errTime = err.Type == ArgError
|
|
if !errTime {
|
|
timeValue = (float64(h)*3600 + float64(m)*60 + s) / 86400
|
|
}
|
|
}
|
|
y, m, d, _, err := strToDate(text)
|
|
errDate = err.Type == ArgError
|
|
if !errDate {
|
|
dateValue = daysBetween(excelMinTime1900.Unix(), makeDate(y, time.Month(m), d)) + 1
|
|
}
|
|
if errTime && errDate {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newNumberFormulaArg(dateValue + timeValue)
|
|
}
|
|
|
|
// Conditional Functions
|
|
|
|
// IF function tests a supplied condition and returns one result if the
|
|
// condition evaluates to TRUE, and another result if the condition evaluates
|
|
// to FALSE. The syntax of the function is:
|
|
//
|
|
// IF(logical_test,value_if_true,value_if_false)
|
|
func (fn *formulaFuncs) IF(argsList *list.List) formulaArg {
|
|
if argsList.Len() == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IF requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IF accepts at most 3 arguments")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg)
|
|
var (
|
|
cond bool
|
|
err error
|
|
result formulaArg
|
|
)
|
|
switch token.Type {
|
|
case ArgString:
|
|
if cond, err = strconv.ParseBool(token.String); err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, err.Error())
|
|
}
|
|
case ArgNumber:
|
|
cond = token.Number == 1
|
|
}
|
|
|
|
if argsList.Len() == 1 {
|
|
return newBoolFormulaArg(cond)
|
|
}
|
|
if cond {
|
|
value := argsList.Front().Next().Value.(formulaArg)
|
|
switch value.Type {
|
|
case ArgNumber:
|
|
result = value.ToNumber()
|
|
default:
|
|
result = newStringFormulaArg(value.String)
|
|
}
|
|
return result
|
|
}
|
|
if argsList.Len() == 3 {
|
|
value := argsList.Back().Value.(formulaArg)
|
|
switch value.Type {
|
|
case ArgNumber:
|
|
result = value.ToNumber()
|
|
default:
|
|
result = newStringFormulaArg(value.String)
|
|
}
|
|
}
|
|
return result
|
|
}
|
|
|
|
// Lookup and Reference Functions
|
|
|
|
// ADDRESS function takes a row and a column number and returns a cell
|
|
// reference as a text string. The syntax of the function is:
|
|
//
|
|
// ADDRESS(row_num,column_num,[abs_num],[a1],[sheet_text])
|
|
func (fn *formulaFuncs) ADDRESS(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ADDRESS requires at least 2 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ADDRESS requires at most 5 arguments")
|
|
}
|
|
rowNum := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rowNum.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if rowNum.Number >= TotalRows {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
colNum := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if colNum.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
absNum := newNumberFormulaArg(1)
|
|
if argsList.Len() >= 3 {
|
|
absNum = argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if absNum.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
if absNum.Number < 1 || absNum.Number > 4 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
a1 := newBoolFormulaArg(true)
|
|
if argsList.Len() >= 4 {
|
|
a1 = argsList.Front().Next().Next().Next().Value.(formulaArg).ToBool()
|
|
if a1.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
var sheetText string
|
|
if argsList.Len() == 5 {
|
|
sheetText = fmt.Sprintf("%s!", argsList.Back().Value.(formulaArg).Value())
|
|
}
|
|
formatter := addressFmtMaps[fmt.Sprintf("%d_%s", int(absNum.Number), a1.Value())]
|
|
addr, err := formatter(int(colNum.Number), int(rowNum.Number))
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newStringFormulaArg(fmt.Sprintf("%s%s", sheetText, addr))
|
|
}
|
|
|
|
// CHOOSE function returns a value from an array, that corresponds to a
|
|
// supplied index number (position). The syntax of the function is:
|
|
//
|
|
// CHOOSE(index_num,value1,[value2],...)
|
|
func (fn *formulaFuncs) CHOOSE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHOOSE requires 2 arguments")
|
|
}
|
|
idx, err := strconv.Atoi(argsList.Front().Value.(formulaArg).Value())
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "CHOOSE requires first argument of type number")
|
|
}
|
|
if argsList.Len() <= idx {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "index_num should be <= to the number of values")
|
|
}
|
|
arg := argsList.Front()
|
|
for i := 0; i < idx; i++ {
|
|
arg = arg.Next()
|
|
}
|
|
return arg.Value.(formulaArg)
|
|
}
|
|
|
|
// deepMatchRune finds whether the text deep matches/satisfies the pattern
|
|
// string.
|
|
func deepMatchRune(str, pattern []rune, simple bool) bool {
|
|
for len(pattern) > 0 {
|
|
switch pattern[0] {
|
|
default:
|
|
if len(str) == 0 || str[0] != pattern[0] {
|
|
return false
|
|
}
|
|
case '?':
|
|
if len(str) == 0 && !simple {
|
|
return false
|
|
}
|
|
case '*':
|
|
return deepMatchRune(str, pattern[1:], simple) ||
|
|
(len(str) > 0 && deepMatchRune(str[1:], pattern, simple))
|
|
}
|
|
str = str[1:]
|
|
pattern = pattern[1:]
|
|
}
|
|
return len(str) == 0 && len(pattern) == 0
|
|
}
|
|
|
|
// matchPattern finds whether the text matches or satisfies the pattern
|
|
// string. The pattern supports '*' and '?' wildcards in the pattern string.
|
|
func matchPattern(pattern, name string) (matched bool) {
|
|
if pattern == "" {
|
|
return name == pattern
|
|
}
|
|
if pattern == "*" {
|
|
return true
|
|
}
|
|
rName, rPattern := make([]rune, 0, len(name)), make([]rune, 0, len(pattern))
|
|
for _, r := range name {
|
|
rName = append(rName, r)
|
|
}
|
|
for _, r := range pattern {
|
|
rPattern = append(rPattern, r)
|
|
}
|
|
return deepMatchRune(rName, rPattern, false)
|
|
}
|
|
|
|
// compareFormulaArg compares the left-hand sides and the right-hand sides'
|
|
// formula arguments by given conditions such as case-sensitive, if exact
|
|
// match, and make compare result as formula criteria condition type.
|
|
func compareFormulaArg(lhs, rhs, matchMode formulaArg, caseSensitive bool) byte {
|
|
if lhs.Type != rhs.Type {
|
|
return criteriaNe
|
|
}
|
|
switch lhs.Type {
|
|
case ArgNumber:
|
|
if lhs.Number == rhs.Number {
|
|
return criteriaEq
|
|
}
|
|
if lhs.Number < rhs.Number {
|
|
return criteriaL
|
|
}
|
|
return criteriaG
|
|
case ArgString:
|
|
ls, rs := lhs.String, rhs.String
|
|
if !caseSensitive {
|
|
ls, rs = strings.ToLower(ls), strings.ToLower(rs)
|
|
}
|
|
if matchMode.Number == matchModeWildcard {
|
|
if matchPattern(rs, ls) {
|
|
return criteriaEq
|
|
}
|
|
}
|
|
return map[int]byte{1: criteriaG, -1: criteriaL, 0: criteriaEq}[strings.Compare(ls, rs)]
|
|
case ArgEmpty:
|
|
return criteriaEq
|
|
case ArgList:
|
|
return compareFormulaArgList(lhs, rhs, matchMode, caseSensitive)
|
|
case ArgMatrix:
|
|
return compareFormulaArgMatrix(lhs, rhs, matchMode, caseSensitive)
|
|
default:
|
|
return criteriaErr
|
|
}
|
|
}
|
|
|
|
// compareFormulaArgList compares the left-hand sides and the right-hand sides
|
|
// list type formula arguments.
|
|
func compareFormulaArgList(lhs, rhs, matchMode formulaArg, caseSensitive bool) byte {
|
|
if len(lhs.List) < len(rhs.List) {
|
|
return criteriaL
|
|
}
|
|
if len(lhs.List) > len(rhs.List) {
|
|
return criteriaG
|
|
}
|
|
for arg := range lhs.List {
|
|
criteria := compareFormulaArg(lhs.List[arg], rhs.List[arg], matchMode, caseSensitive)
|
|
if criteria != criteriaEq {
|
|
return criteria
|
|
}
|
|
}
|
|
return criteriaEq
|
|
}
|
|
|
|
// compareFormulaArgMatrix compares the left-hand sides and the right-hand sides'
|
|
// matrix type formula arguments.
|
|
func compareFormulaArgMatrix(lhs, rhs, matchMode formulaArg, caseSensitive bool) byte {
|
|
if len(lhs.Matrix) < len(rhs.Matrix) {
|
|
return criteriaL
|
|
}
|
|
if len(lhs.Matrix) > len(rhs.Matrix) {
|
|
return criteriaG
|
|
}
|
|
for i := range lhs.Matrix {
|
|
left := lhs.Matrix[i]
|
|
right := lhs.Matrix[i]
|
|
if len(left) < len(right) {
|
|
return criteriaL
|
|
}
|
|
if len(left) > len(right) {
|
|
return criteriaG
|
|
}
|
|
for arg := range left {
|
|
criteria := compareFormulaArg(left[arg], right[arg], matchMode, caseSensitive)
|
|
if criteria != criteriaEq {
|
|
return criteria
|
|
}
|
|
}
|
|
}
|
|
return criteriaEq
|
|
}
|
|
|
|
// COLUMN function returns the first column number within a supplied reference
|
|
// or the number of the current column. The syntax of the function is:
|
|
//
|
|
// COLUMN([reference])
|
|
func (fn *formulaFuncs) COLUMN(argsList *list.List) formulaArg {
|
|
if argsList.Len() > 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COLUMN requires at most 1 argument")
|
|
}
|
|
if argsList.Len() == 1 {
|
|
if argsList.Front().Value.(formulaArg).cellRanges != nil && argsList.Front().Value.(formulaArg).cellRanges.Len() > 0 {
|
|
return newNumberFormulaArg(float64(argsList.Front().Value.(formulaArg).cellRanges.Front().Value.(cellRange).From.Col))
|
|
}
|
|
if argsList.Front().Value.(formulaArg).cellRefs != nil && argsList.Front().Value.(formulaArg).cellRefs.Len() > 0 {
|
|
return newNumberFormulaArg(float64(argsList.Front().Value.(formulaArg).cellRefs.Front().Value.(cellRef).Col))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, "invalid reference")
|
|
}
|
|
col, _, _ := CellNameToCoordinates(fn.cell)
|
|
return newNumberFormulaArg(float64(col))
|
|
}
|
|
|
|
// calcColumnsMinMax calculation min and max value for given formula arguments
|
|
// sequence of the formula function COLUMNS.
|
|
func calcColumnsMinMax(argsList *list.List) (min, max int) {
|
|
if argsList.Front().Value.(formulaArg).cellRanges != nil && argsList.Front().Value.(formulaArg).cellRanges.Len() > 0 {
|
|
crs := argsList.Front().Value.(formulaArg).cellRanges
|
|
for cr := crs.Front(); cr != nil; cr = cr.Next() {
|
|
if min == 0 {
|
|
min = cr.Value.(cellRange).From.Col
|
|
}
|
|
if min > cr.Value.(cellRange).From.Col {
|
|
min = cr.Value.(cellRange).From.Col
|
|
}
|
|
if min > cr.Value.(cellRange).To.Col {
|
|
min = cr.Value.(cellRange).To.Col
|
|
}
|
|
if max < cr.Value.(cellRange).To.Col {
|
|
max = cr.Value.(cellRange).To.Col
|
|
}
|
|
if max < cr.Value.(cellRange).From.Col {
|
|
max = cr.Value.(cellRange).From.Col
|
|
}
|
|
}
|
|
}
|
|
if argsList.Front().Value.(formulaArg).cellRefs != nil && argsList.Front().Value.(formulaArg).cellRefs.Len() > 0 {
|
|
cr := argsList.Front().Value.(formulaArg).cellRefs
|
|
for refs := cr.Front(); refs != nil; refs = refs.Next() {
|
|
if min == 0 {
|
|
min = refs.Value.(cellRef).Col
|
|
}
|
|
if min > refs.Value.(cellRef).Col {
|
|
min = refs.Value.(cellRef).Col
|
|
}
|
|
if max < refs.Value.(cellRef).Col {
|
|
max = refs.Value.(cellRef).Col
|
|
}
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// COLUMNS function receives an Excel range and returns the number of columns
|
|
// that are contained within the range. The syntax of the function is:
|
|
//
|
|
// COLUMNS(array)
|
|
func (fn *formulaFuncs) COLUMNS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "COLUMNS requires 1 argument")
|
|
}
|
|
min, max := calcColumnsMinMax(argsList)
|
|
if max == MaxColumns {
|
|
return newNumberFormulaArg(float64(MaxColumns))
|
|
}
|
|
result := max - min + 1
|
|
if max == min {
|
|
if min == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "invalid reference")
|
|
}
|
|
return newNumberFormulaArg(float64(1))
|
|
}
|
|
return newNumberFormulaArg(float64(result))
|
|
}
|
|
|
|
// FORMULATEXT function returns a formula as a text string. The syntax of the
|
|
// function is:
|
|
//
|
|
// FORMULATEXT(reference)
|
|
func (fn *formulaFuncs) FORMULATEXT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FORMULATEXT requires 1 argument")
|
|
}
|
|
refs := argsList.Front().Value.(formulaArg).cellRefs
|
|
col, row := 0, 0
|
|
if refs != nil && refs.Len() > 0 {
|
|
col, row = refs.Front().Value.(cellRef).Col, refs.Front().Value.(cellRef).Row
|
|
}
|
|
ranges := argsList.Front().Value.(formulaArg).cellRanges
|
|
if ranges != nil && ranges.Len() > 0 {
|
|
col, row = ranges.Front().Value.(cellRange).From.Col, ranges.Front().Value.(cellRange).From.Row
|
|
}
|
|
cell, err := CoordinatesToCellName(col, row)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
formula, _ := fn.f.GetCellFormula(fn.sheet, cell)
|
|
return newStringFormulaArg(formula)
|
|
}
|
|
|
|
// checkHVLookupArgs checking arguments, prepare extract mode, lookup value,
|
|
// and data for the formula functions HLOOKUP and VLOOKUP.
|
|
func checkHVLookupArgs(name string, argsList *list.List) (idx int, lookupValue, tableArray, matchMode, errArg formulaArg) {
|
|
unit := map[string]string{
|
|
"HLOOKUP": "row",
|
|
"VLOOKUP": "col",
|
|
}[name]
|
|
if argsList.Len() < 3 {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 3 arguments", name))
|
|
return
|
|
}
|
|
if argsList.Len() > 4 {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at most 4 arguments", name))
|
|
return
|
|
}
|
|
lookupValue = argsList.Front().Value.(formulaArg)
|
|
tableArray = argsList.Front().Next().Value.(formulaArg)
|
|
if tableArray.Type != ArgMatrix {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires second argument of table array", name))
|
|
return
|
|
}
|
|
arg := argsList.Front().Next().Next().Value.(formulaArg)
|
|
if arg.Type != ArgNumber || arg.Boolean {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires numeric %s argument", name, unit))
|
|
return
|
|
}
|
|
idx, matchMode = int(arg.Number)-1, newNumberFormulaArg(matchModeMaxLess)
|
|
if argsList.Len() == 4 {
|
|
rangeLookup := argsList.Back().Value.(formulaArg).ToBool()
|
|
if rangeLookup.Type == ArgError {
|
|
errArg = rangeLookup
|
|
return
|
|
}
|
|
if rangeLookup.Number == 0 {
|
|
matchMode = newNumberFormulaArg(matchModeWildcard)
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// HLOOKUP function 'looks up' a given value in the top row of a data array
|
|
// (or table), and returns the corresponding value from another row of the
|
|
// array. The syntax of the function is:
|
|
//
|
|
// HLOOKUP(lookup_value,table_array,row_index_num,[range_lookup])
|
|
func (fn *formulaFuncs) HLOOKUP(argsList *list.List) formulaArg {
|
|
rowIdx, lookupValue, tableArray, matchMode, errArg := checkHVLookupArgs("HLOOKUP", argsList)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
var matchIdx int
|
|
var wasExact bool
|
|
if matchMode.Number == matchModeWildcard || len(tableArray.Matrix) == TotalRows {
|
|
matchIdx, wasExact = lookupLinearSearch(false, lookupValue, tableArray, matchMode, newNumberFormulaArg(searchModeLinear))
|
|
} else {
|
|
matchIdx, wasExact = lookupBinarySearch(false, lookupValue, tableArray, matchMode, newNumberFormulaArg(searchModeAscBinary))
|
|
}
|
|
if matchIdx == -1 {
|
|
return newErrorFormulaArg(formulaErrorNA, "HLOOKUP no result found")
|
|
}
|
|
if rowIdx < 0 || rowIdx >= len(tableArray.Matrix) {
|
|
return newErrorFormulaArg(formulaErrorNA, "HLOOKUP has invalid row index")
|
|
}
|
|
row := tableArray.Matrix[rowIdx]
|
|
if wasExact || matchMode.Number == matchModeWildcard {
|
|
return row[matchIdx]
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, "HLOOKUP no result found")
|
|
}
|
|
|
|
// HYPERLINK function creates a hyperlink to a specified location. The syntax
|
|
// of the function is:
|
|
//
|
|
// HYPERLINK(link_location,[friendly_name])
|
|
func (fn *formulaFuncs) HYPERLINK(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HYPERLINK requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "HYPERLINK allows at most 2 arguments")
|
|
}
|
|
return newStringFormulaArg(argsList.Back().Value.(formulaArg).Value())
|
|
}
|
|
|
|
// calcMatch returns the position of the value by given match type, criteria
|
|
// and lookup array for the formula function MATCH.
|
|
func calcMatch(matchType int, criteria *formulaCriteria, lookupArray []formulaArg) formulaArg {
|
|
switch matchType {
|
|
case 0:
|
|
for i, arg := range lookupArray {
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), criteria); ok {
|
|
return newNumberFormulaArg(float64(i + 1))
|
|
}
|
|
}
|
|
case -1:
|
|
for i, arg := range lookupArray {
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), criteria); ok {
|
|
return newNumberFormulaArg(float64(i + 1))
|
|
}
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), &formulaCriteria{
|
|
Type: criteriaL, Condition: criteria.Condition,
|
|
}); ok {
|
|
if i == 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(float64(i))
|
|
}
|
|
}
|
|
case 1:
|
|
for i, arg := range lookupArray {
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), criteria); ok {
|
|
return newNumberFormulaArg(float64(i + 1))
|
|
}
|
|
if ok, _ := formulaCriteriaEval(arg.Value(), &formulaCriteria{
|
|
Type: criteriaG, Condition: criteria.Condition,
|
|
}); ok {
|
|
if i == 0 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
return newNumberFormulaArg(float64(i))
|
|
}
|
|
}
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
|
|
// MATCH function looks up a value in an array, and returns the position of
|
|
// the value within the array. The user can specify that the function should
|
|
// only return a result if an exact match is found, or that the function
|
|
// should return the position of the closest match (above or below), if an
|
|
// exact match is not found. The syntax of the Match function is:
|
|
//
|
|
// MATCH(lookup_value,lookup_array,[match_type])
|
|
func (fn *formulaFuncs) MATCH(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 && argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MATCH requires 1 or 2 arguments")
|
|
}
|
|
var (
|
|
matchType = 1
|
|
lookupArray []formulaArg
|
|
lookupArrayArg = argsList.Front().Next().Value.(formulaArg)
|
|
lookupArrayErr = "MATCH arguments lookup_array should be one-dimensional array"
|
|
)
|
|
if argsList.Len() == 3 {
|
|
matchTypeArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if matchTypeArg.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MATCH requires numeric match_type argument")
|
|
}
|
|
if matchTypeArg.Number == -1 || matchTypeArg.Number == 0 {
|
|
matchType = int(matchTypeArg.Number)
|
|
}
|
|
}
|
|
switch lookupArrayArg.Type {
|
|
case ArgMatrix:
|
|
if len(lookupArrayArg.Matrix[0]) != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, lookupArrayErr)
|
|
}
|
|
lookupArray = lookupArrayArg.ToList()
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorNA, lookupArrayErr)
|
|
}
|
|
return calcMatch(matchType, formulaCriteriaParser(argsList.Front().Value.(formulaArg).Value()), lookupArray)
|
|
}
|
|
|
|
// TRANSPOSE function 'transposes' an array of cells (i.e. the function copies
|
|
// a horizontal range of cells into a vertical range and vice versa). The
|
|
// syntax of the function is:
|
|
//
|
|
// TRANSPOSE(array)
|
|
func (fn *formulaFuncs) TRANSPOSE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TRANSPOSE requires 1 argument")
|
|
}
|
|
args := argsList.Back().Value.(formulaArg).ToList()
|
|
rmin, rmax := calcRowsMinMax(argsList)
|
|
cmin, cmax := calcColumnsMinMax(argsList)
|
|
cols, rows := cmax-cmin+1, rmax-rmin+1
|
|
src := make([][]formulaArg, 0)
|
|
for i := 0; i < len(args); i += cols {
|
|
src = append(src, args[i:i+cols])
|
|
}
|
|
mtx := make([][]formulaArg, cols)
|
|
for r, row := range src {
|
|
colIdx := r % rows
|
|
for c, cell := range row {
|
|
rowIdx := c % cols
|
|
if len(mtx[rowIdx]) == 0 {
|
|
mtx[rowIdx] = make([]formulaArg, rows)
|
|
}
|
|
mtx[rowIdx][colIdx] = cell
|
|
}
|
|
}
|
|
return newMatrixFormulaArg(mtx)
|
|
}
|
|
|
|
// lookupLinearSearch sequentially checks each look value of the lookup array until
|
|
// a match is found or the whole list has been searched.
|
|
func lookupLinearSearch(vertical bool, lookupValue, lookupArray, matchMode, searchMode formulaArg) (int, bool) {
|
|
var tableArray []formulaArg
|
|
if vertical {
|
|
for _, row := range lookupArray.Matrix {
|
|
tableArray = append(tableArray, row[0])
|
|
}
|
|
} else {
|
|
tableArray = lookupArray.Matrix[0]
|
|
}
|
|
matchIdx, wasExact := -1, false
|
|
start:
|
|
for i, cell := range tableArray {
|
|
lhs := cell
|
|
if lookupValue.Type == ArgNumber {
|
|
if lhs = cell.ToNumber(); lhs.Type == ArgError {
|
|
lhs = cell
|
|
}
|
|
} else if lookupValue.Type == ArgMatrix {
|
|
lhs = lookupArray
|
|
} else if lookupArray.Type == ArgString {
|
|
lhs = newStringFormulaArg(cell.Value())
|
|
}
|
|
if compareFormulaArg(lhs, lookupValue, matchMode, false) == criteriaEq {
|
|
matchIdx = i
|
|
wasExact = true
|
|
if searchMode.Number == searchModeLinear {
|
|
break start
|
|
}
|
|
}
|
|
if matchMode.Number == matchModeMinGreater || matchMode.Number == matchModeMaxLess {
|
|
matchIdx = int(calcMatch(int(matchMode.Number), formulaCriteriaParser(lookupValue.Value()), tableArray).Number)
|
|
continue
|
|
}
|
|
}
|
|
return matchIdx, wasExact
|
|
}
|
|
|
|
// VLOOKUP function 'looks up' a given value in the left-hand column of a
|
|
// data array (or table), and returns the corresponding value from another
|
|
// column of the array. The syntax of the function is:
|
|
//
|
|
// VLOOKUP(lookup_value,table_array,col_index_num,[range_lookup])
|
|
func (fn *formulaFuncs) VLOOKUP(argsList *list.List) formulaArg {
|
|
colIdx, lookupValue, tableArray, matchMode, errArg := checkHVLookupArgs("VLOOKUP", argsList)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
var matchIdx int
|
|
var wasExact bool
|
|
if matchMode.Number == matchModeWildcard || len(tableArray.Matrix) == TotalRows {
|
|
matchIdx, wasExact = lookupLinearSearch(true, lookupValue, tableArray, matchMode, newNumberFormulaArg(searchModeLinear))
|
|
} else {
|
|
matchIdx, wasExact = lookupBinarySearch(true, lookupValue, tableArray, matchMode, newNumberFormulaArg(searchModeAscBinary))
|
|
}
|
|
if matchIdx == -1 {
|
|
return newErrorFormulaArg(formulaErrorNA, "VLOOKUP no result found")
|
|
}
|
|
mtx := tableArray.Matrix[matchIdx]
|
|
if colIdx < 0 || colIdx >= len(mtx) {
|
|
return newErrorFormulaArg(formulaErrorNA, "VLOOKUP has invalid column index")
|
|
}
|
|
if wasExact || matchMode.Number == matchModeWildcard {
|
|
return mtx[colIdx]
|
|
}
|
|
return newErrorFormulaArg(formulaErrorNA, "VLOOKUP no result found")
|
|
}
|
|
|
|
// lookupBinarySearch finds the position of a target value when range lookup
|
|
// is TRUE, if the data of table array can't guarantee be sorted, it will
|
|
// return wrong result.
|
|
func lookupBinarySearch(vertical bool, lookupValue, lookupArray, matchMode, searchMode formulaArg) (matchIdx int, wasExact bool) {
|
|
var tableArray []formulaArg
|
|
if vertical {
|
|
for _, row := range lookupArray.Matrix {
|
|
tableArray = append(tableArray, row[0])
|
|
}
|
|
} else {
|
|
tableArray = lookupArray.Matrix[0]
|
|
}
|
|
low, high, lastMatchIdx := 0, len(tableArray)-1, -1
|
|
count := high
|
|
for low <= high {
|
|
mid := low + (high-low)/2
|
|
cell := tableArray[mid]
|
|
lhs := cell
|
|
if lookupValue.Type == ArgNumber {
|
|
if lhs = cell.ToNumber(); lhs.Type == ArgError {
|
|
lhs = cell
|
|
}
|
|
} else if lookupValue.Type == ArgMatrix && vertical {
|
|
lhs = lookupArray
|
|
} else if lookupValue.Type == ArgString {
|
|
lhs = newStringFormulaArg(cell.Value())
|
|
}
|
|
result := compareFormulaArg(lhs, lookupValue, matchMode, false)
|
|
if result == criteriaEq {
|
|
matchIdx, wasExact = mid, true
|
|
if searchMode.Number == searchModeDescBinary {
|
|
matchIdx = count - matchIdx
|
|
}
|
|
return
|
|
} else if result == criteriaG {
|
|
high = mid - 1
|
|
} else if result == criteriaL {
|
|
matchIdx = mid
|
|
if cell.Type != ArgEmpty {
|
|
lastMatchIdx = matchIdx
|
|
}
|
|
low = mid + 1
|
|
} else {
|
|
return -1, false
|
|
}
|
|
}
|
|
matchIdx, wasExact = lastMatchIdx, true
|
|
return
|
|
}
|
|
|
|
// checkLookupArgs checking arguments, prepare lookup value, and data for the
|
|
// formula function LOOKUP.
|
|
func checkLookupArgs(argsList *list.List) (arrayForm bool, lookupValue, lookupVector, errArg formulaArg) {
|
|
if argsList.Len() < 2 {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, "LOOKUP requires at least 2 arguments")
|
|
return
|
|
}
|
|
if argsList.Len() > 3 {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, "LOOKUP requires at most 3 arguments")
|
|
return
|
|
}
|
|
lookupValue = newStringFormulaArg(argsList.Front().Value.(formulaArg).Value())
|
|
lookupVector = argsList.Front().Next().Value.(formulaArg)
|
|
if lookupVector.Type != ArgMatrix && lookupVector.Type != ArgList {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, "LOOKUP requires second argument of table array")
|
|
return
|
|
}
|
|
arrayForm = lookupVector.Type == ArgMatrix
|
|
if arrayForm && len(lookupVector.Matrix) == 0 {
|
|
errArg = newErrorFormulaArg(formulaErrorVALUE, "LOOKUP requires not empty range as second argument")
|
|
}
|
|
return
|
|
}
|
|
|
|
// iterateLookupArgs iterate arguments to extract columns and calculate match
|
|
// index for the formula function LOOKUP.
|
|
func iterateLookupArgs(lookupValue, lookupVector formulaArg) ([]formulaArg, int, bool) {
|
|
cols, matchIdx, ok := lookupCol(lookupVector, 0), -1, false
|
|
for idx, col := range cols {
|
|
lhs := lookupValue
|
|
switch col.Type {
|
|
case ArgNumber:
|
|
lhs = lhs.ToNumber()
|
|
if !col.Boolean {
|
|
if lhs.Type == ArgError {
|
|
lhs = lookupValue
|
|
}
|
|
}
|
|
}
|
|
compare := compareFormulaArg(lhs, col, newNumberFormulaArg(matchModeMaxLess), false)
|
|
// Find exact match
|
|
if compare == criteriaEq {
|
|
matchIdx = idx
|
|
break
|
|
}
|
|
// Find the nearest match if lookup value is more than or equal to the first value in lookup vector
|
|
if idx == 0 {
|
|
ok = compare == criteriaG
|
|
} else if ok && compare == criteriaL && matchIdx == -1 {
|
|
matchIdx = idx - 1
|
|
}
|
|
}
|
|
return cols, matchIdx, ok
|
|
}
|
|
|
|
// index is an implementation of the formula function INDEX.
|
|
func (fn *formulaFuncs) index(array formulaArg, rowIdx, colIdx int) formulaArg {
|
|
var cells []formulaArg
|
|
if array.Type == ArgMatrix {
|
|
cellMatrix := array.Matrix
|
|
if rowIdx < -1 || rowIdx >= len(cellMatrix) {
|
|
return newErrorFormulaArg(formulaErrorREF, "INDEX row_num out of range")
|
|
}
|
|
if rowIdx == -1 {
|
|
if colIdx >= len(cellMatrix[0]) {
|
|
return newErrorFormulaArg(formulaErrorREF, "INDEX col_num out of range")
|
|
}
|
|
var column [][]formulaArg
|
|
for _, cells = range cellMatrix {
|
|
column = append(column, []formulaArg{cells[colIdx]})
|
|
}
|
|
return newMatrixFormulaArg(column)
|
|
}
|
|
cells = cellMatrix[rowIdx]
|
|
}
|
|
if colIdx < -1 || colIdx >= len(cells) {
|
|
return newErrorFormulaArg(formulaErrorREF, "INDEX col_num out of range")
|
|
}
|
|
return newListFormulaArg(cells)
|
|
}
|
|
|
|
// validateMatchMode check the number of match mode if be equal to 0, 1, -1 or
|
|
// 2.
|
|
func validateMatchMode(mode float64) bool {
|
|
return mode == matchModeExact || mode == matchModeMinGreater || mode == matchModeMaxLess || mode == matchModeWildcard
|
|
}
|
|
|
|
// validateSearchMode check the number of search mode if be equal to 1, -1, 2
|
|
// or -2.
|
|
func validateSearchMode(mode float64) bool {
|
|
return mode == searchModeLinear || mode == searchModeReverseLinear || mode == searchModeAscBinary || mode == searchModeDescBinary
|
|
}
|
|
|
|
// prepareXlookupArgs checking and prepare arguments for the formula function
|
|
// XLOOKUP.
|
|
func (fn *formulaFuncs) prepareXlookupArgs(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XLOOKUP requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XLOOKUP allows at most 6 arguments")
|
|
}
|
|
lookupValue := argsList.Front().Value.(formulaArg)
|
|
lookupArray := argsList.Front().Next().Value.(formulaArg)
|
|
returnArray := argsList.Front().Next().Next().Value.(formulaArg)
|
|
ifNotFond := newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
matchMode, searchMode := newNumberFormulaArg(matchModeExact), newNumberFormulaArg(searchModeLinear)
|
|
if argsList.Len() > 3 {
|
|
ifNotFond = argsList.Front().Next().Next().Next().Value.(formulaArg)
|
|
}
|
|
if argsList.Len() > 4 {
|
|
if matchMode = argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); matchMode.Type != ArgNumber {
|
|
return matchMode
|
|
}
|
|
}
|
|
if argsList.Len() > 5 {
|
|
if searchMode = argsList.Back().Value.(formulaArg).ToNumber(); searchMode.Type != ArgNumber {
|
|
return searchMode
|
|
}
|
|
}
|
|
if lookupArray.Type != ArgMatrix || returnArray.Type != ArgMatrix {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if !validateMatchMode(matchMode.Number) || !validateSearchMode(searchMode.Number) {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return newListFormulaArg([]formulaArg{lookupValue, lookupArray, returnArray, ifNotFond, matchMode, searchMode})
|
|
}
|
|
|
|
// xlookup is an implementation of the formula function XLOOKUP.
|
|
func (fn *formulaFuncs) xlookup(lookupRows, lookupCols, returnArrayRows, returnArrayCols, matchIdx int,
|
|
condition1, condition2, condition3, condition4 bool, returnArray formulaArg,
|
|
) formulaArg {
|
|
var result [][]formulaArg
|
|
for rowIdx, row := range returnArray.Matrix {
|
|
for colIdx, cell := range row {
|
|
if condition1 {
|
|
if condition2 {
|
|
result = append(result, []formulaArg{cell})
|
|
continue
|
|
}
|
|
if returnArrayRows > 1 && returnArrayCols > 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
if condition3 {
|
|
if returnArrayCols != lookupCols {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if colIdx == matchIdx {
|
|
result = append(result, []formulaArg{cell})
|
|
continue
|
|
}
|
|
}
|
|
if condition4 {
|
|
if returnArrayRows != lookupRows {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if rowIdx == matchIdx {
|
|
if len(result) == 0 {
|
|
result = append(result, []formulaArg{cell})
|
|
continue
|
|
}
|
|
result[0] = append(result[0], cell)
|
|
}
|
|
}
|
|
}
|
|
}
|
|
array := newMatrixFormulaArg(result)
|
|
cells := array.ToList()
|
|
if len(cells) == 1 {
|
|
return cells[0]
|
|
}
|
|
return array
|
|
}
|
|
|
|
// XLOOKUP function searches a range or an array, and then returns the item
|
|
// corresponding to the first match it finds. If no match exists, then
|
|
// XLOOKUP can return the closest (approximate) match. The syntax of the
|
|
// function is:
|
|
//
|
|
// XLOOKUP(lookup_value,lookup_array,return_array,[if_not_found],[match_mode],[search_mode])
|
|
func (fn *formulaFuncs) XLOOKUP(argsList *list.List) formulaArg {
|
|
args := fn.prepareXlookupArgs(argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
lookupValue, lookupArray, returnArray, ifNotFond, matchMode, searchMode := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5]
|
|
lookupRows, lookupCols := len(lookupArray.Matrix), 0
|
|
if lookupRows > 0 {
|
|
lookupCols = len(lookupArray.Matrix[0])
|
|
}
|
|
if lookupRows != 1 && lookupCols != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
verticalLookup := lookupRows >= lookupCols
|
|
var matchIdx int
|
|
switch searchMode.Number {
|
|
case searchModeLinear, searchModeReverseLinear:
|
|
matchIdx, _ = lookupLinearSearch(verticalLookup, lookupValue, lookupArray, matchMode, searchMode)
|
|
default:
|
|
matchIdx, _ = lookupBinarySearch(verticalLookup, lookupValue, lookupArray, matchMode, searchMode)
|
|
}
|
|
if matchIdx == -1 {
|
|
return ifNotFond
|
|
}
|
|
returnArrayRows, returnArrayCols := len(returnArray.Matrix), len(returnArray.Matrix[0])
|
|
condition1 := lookupRows == 1 && lookupCols == 1
|
|
condition2 := returnArrayRows == 1 || returnArrayCols == 1
|
|
condition3 := lookupRows == 1 && lookupCols > 1
|
|
condition4 := lookupRows > 1 && lookupCols == 1
|
|
return fn.xlookup(lookupRows, lookupCols, returnArrayRows, returnArrayCols, matchIdx, condition1, condition2, condition3, condition4, returnArray)
|
|
}
|
|
|
|
// INDEX function returns a reference to a cell that lies in a specified row
|
|
// and column of a range of cells. The syntax of the function is:
|
|
//
|
|
// INDEX(array,row_num,[col_num])
|
|
func (fn *formulaFuncs) INDEX(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 || argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "INDEX requires 2 or 3 arguments")
|
|
}
|
|
array := argsList.Front().Value.(formulaArg)
|
|
if array.Type != ArgMatrix && array.Type != ArgList {
|
|
array = newMatrixFormulaArg([][]formulaArg{{array}})
|
|
}
|
|
rowArg := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if rowArg.Type != ArgNumber {
|
|
return rowArg
|
|
}
|
|
rowIdx, colIdx := int(rowArg.Number)-1, -1
|
|
if argsList.Len() == 3 {
|
|
colArg := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if colArg.Type != ArgNumber {
|
|
return colArg
|
|
}
|
|
colIdx = int(colArg.Number) - 1
|
|
}
|
|
if rowIdx == -1 && colIdx == -1 {
|
|
if len(array.ToList()) != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return array.ToList()[0]
|
|
}
|
|
cells := fn.index(array, rowIdx, colIdx)
|
|
if cells.Type != ArgList {
|
|
return cells
|
|
}
|
|
if colIdx == -1 {
|
|
return newMatrixFormulaArg([][]formulaArg{cells.List})
|
|
}
|
|
return cells.List[colIdx]
|
|
}
|
|
|
|
// INDIRECT function converts a text string into a cell reference. The syntax
|
|
// of the Indirect function is:
|
|
//
|
|
// INDIRECT(ref_text,[a1])
|
|
func (fn *formulaFuncs) INDIRECT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 && argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "INDIRECT requires 1 or 2 arguments")
|
|
}
|
|
refText := argsList.Front().Value.(formulaArg).Value()
|
|
a1 := newBoolFormulaArg(true)
|
|
if argsList.Len() == 2 {
|
|
if a1 = argsList.Back().Value.(formulaArg).ToBool(); a1.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
R1C1ToA1 := func(ref string) (cell string, err error) {
|
|
parts := strings.Split(strings.TrimLeft(ref, "R"), "C")
|
|
if len(parts) != 2 {
|
|
return
|
|
}
|
|
row, err := strconv.Atoi(parts[0])
|
|
if err != nil {
|
|
return
|
|
}
|
|
col, err := strconv.Atoi(parts[1])
|
|
if err != nil {
|
|
return
|
|
}
|
|
cell, err = CoordinatesToCellName(col, row)
|
|
return
|
|
}
|
|
refs := strings.Split(refText, ":")
|
|
fromRef, toRef := refs[0], ""
|
|
if len(refs) == 2 {
|
|
toRef = refs[1]
|
|
}
|
|
if a1.Number == 0 {
|
|
from, err := R1C1ToA1(refs[0])
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
}
|
|
fromRef = from
|
|
if len(refs) == 2 {
|
|
to, err := R1C1ToA1(refs[1])
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
}
|
|
toRef = to
|
|
}
|
|
}
|
|
if len(refs) == 1 {
|
|
value, err := fn.f.GetCellValue(fn.sheet, fromRef)
|
|
if err != nil {
|
|
return newErrorFormulaArg(formulaErrorREF, formulaErrorREF)
|
|
}
|
|
return newStringFormulaArg(value)
|
|
}
|
|
arg, _ := fn.f.parseReference(fn.ctx, fn.sheet, fromRef+":"+toRef)
|
|
return arg
|
|
}
|
|
|
|
// LOOKUP function performs an approximate match lookup in a one-column or
|
|
// one-row range, and returns the corresponding value from another one-column
|
|
// or one-row range. The syntax of the function is:
|
|
//
|
|
// LOOKUP(lookup_value,lookup_vector,[result_vector])
|
|
func (fn *formulaFuncs) LOOKUP(argsList *list.List) formulaArg {
|
|
arrayForm, lookupValue, lookupVector, errArg := checkLookupArgs(argsList)
|
|
if errArg.Type == ArgError {
|
|
return errArg
|
|
}
|
|
cols, matchIdx, ok := iterateLookupArgs(lookupValue, lookupVector)
|
|
if ok && matchIdx == -1 {
|
|
matchIdx = len(cols) - 1
|
|
}
|
|
var column []formulaArg
|
|
if argsList.Len() == 3 {
|
|
column = lookupCol(argsList.Back().Value.(formulaArg), 0)
|
|
} else if arrayForm && len(lookupVector.Matrix[0]) > 1 {
|
|
column = lookupCol(lookupVector, 1)
|
|
} else {
|
|
column = cols
|
|
}
|
|
if matchIdx < 0 || matchIdx >= len(column) {
|
|
return newErrorFormulaArg(formulaErrorNA, "LOOKUP no result found")
|
|
}
|
|
return column[matchIdx]
|
|
}
|
|
|
|
// lookupCol extract columns for LOOKUP.
|
|
func lookupCol(arr formulaArg, idx int) []formulaArg {
|
|
col := arr.List
|
|
if arr.Type == ArgMatrix {
|
|
col = nil
|
|
for _, r := range arr.Matrix {
|
|
if len(r) > 0 {
|
|
col = append(col, r[idx])
|
|
continue
|
|
}
|
|
col = append(col, newEmptyFormulaArg())
|
|
}
|
|
}
|
|
return col
|
|
}
|
|
|
|
// ROW function returns the first row number within a supplied reference or
|
|
// the number of the current row. The syntax of the function is:
|
|
//
|
|
// ROW([reference])
|
|
func (fn *formulaFuncs) ROW(argsList *list.List) formulaArg {
|
|
if argsList.Len() > 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROW requires at most 1 argument")
|
|
}
|
|
if argsList.Len() == 1 {
|
|
if argsList.Front().Value.(formulaArg).cellRanges != nil && argsList.Front().Value.(formulaArg).cellRanges.Len() > 0 {
|
|
return newNumberFormulaArg(float64(argsList.Front().Value.(formulaArg).cellRanges.Front().Value.(cellRange).From.Row))
|
|
}
|
|
if argsList.Front().Value.(formulaArg).cellRefs != nil && argsList.Front().Value.(formulaArg).cellRefs.Len() > 0 {
|
|
return newNumberFormulaArg(float64(argsList.Front().Value.(formulaArg).cellRefs.Front().Value.(cellRef).Row))
|
|
}
|
|
return newErrorFormulaArg(formulaErrorVALUE, "invalid reference")
|
|
}
|
|
_, row, _ := CellNameToCoordinates(fn.cell)
|
|
return newNumberFormulaArg(float64(row))
|
|
}
|
|
|
|
// calcRowsMinMax calculation min and max value for given formula arguments
|
|
// sequence of the formula function ROWS.
|
|
func calcRowsMinMax(argsList *list.List) (min, max int) {
|
|
if argsList.Front().Value.(formulaArg).cellRanges != nil && argsList.Front().Value.(formulaArg).cellRanges.Len() > 0 {
|
|
crs := argsList.Front().Value.(formulaArg).cellRanges
|
|
for cr := crs.Front(); cr != nil; cr = cr.Next() {
|
|
if min == 0 {
|
|
min = cr.Value.(cellRange).From.Row
|
|
}
|
|
if min > cr.Value.(cellRange).From.Row {
|
|
min = cr.Value.(cellRange).From.Row
|
|
}
|
|
if min > cr.Value.(cellRange).To.Row {
|
|
min = cr.Value.(cellRange).To.Row
|
|
}
|
|
if max < cr.Value.(cellRange).To.Row {
|
|
max = cr.Value.(cellRange).To.Row
|
|
}
|
|
if max < cr.Value.(cellRange).From.Row {
|
|
max = cr.Value.(cellRange).From.Row
|
|
}
|
|
}
|
|
}
|
|
if argsList.Front().Value.(formulaArg).cellRefs != nil && argsList.Front().Value.(formulaArg).cellRefs.Len() > 0 {
|
|
cr := argsList.Front().Value.(formulaArg).cellRefs
|
|
for refs := cr.Front(); refs != nil; refs = refs.Next() {
|
|
if min == 0 {
|
|
min = refs.Value.(cellRef).Row
|
|
}
|
|
if min > refs.Value.(cellRef).Row {
|
|
min = refs.Value.(cellRef).Row
|
|
}
|
|
if max < refs.Value.(cellRef).Row {
|
|
max = refs.Value.(cellRef).Row
|
|
}
|
|
}
|
|
}
|
|
return
|
|
}
|
|
|
|
// ROWS function takes an Excel range and returns the number of rows that are
|
|
// contained within the range. The syntax of the function is:
|
|
//
|
|
// ROWS(array)
|
|
func (fn *formulaFuncs) ROWS(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ROWS requires 1 argument")
|
|
}
|
|
min, max := calcRowsMinMax(argsList)
|
|
if max == TotalRows {
|
|
return newStringFormulaArg(strconv.Itoa(TotalRows))
|
|
}
|
|
result := max - min + 1
|
|
if max == min {
|
|
if min == 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "invalid reference")
|
|
}
|
|
return newNumberFormulaArg(float64(1))
|
|
}
|
|
return newStringFormulaArg(strconv.Itoa(result))
|
|
}
|
|
|
|
// Web Functions
|
|
|
|
// ENCODEURL function returns a URL-encoded string, replacing certain
|
|
// non-alphanumeric characters with the percentage symbol (%) and a
|
|
// hexadecimal number. The syntax of the function is:
|
|
//
|
|
// ENCODEURL(url)
|
|
func (fn *formulaFuncs) ENCODEURL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ENCODEURL requires 1 argument")
|
|
}
|
|
token := argsList.Front().Value.(formulaArg).Value()
|
|
return newStringFormulaArg(strings.ReplaceAll(url.QueryEscape(token), "+", "%20"))
|
|
}
|
|
|
|
// Financial Functions
|
|
|
|
// validateFrequency check the number of coupon payments per year if be equal to 1, 2 or 4.
|
|
func validateFrequency(freq float64) bool {
|
|
return freq == 1 || freq == 2 || freq == 4
|
|
}
|
|
|
|
// ACCRINT function returns the accrued interest in a security that pays
|
|
// periodic interest. The syntax of the function is:
|
|
//
|
|
// ACCRINT(issue,first_interest,settlement,rate,par,frequency,[basis],[calc_method])
|
|
func (fn *formulaFuncs) ACCRINT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACCRINT requires at least 6 arguments")
|
|
}
|
|
if argsList.Len() > 8 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACCRINT allows at most 8 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(3, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
issue, settlement := args.List[0], args.List[2]
|
|
rate := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
par := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
frequency := argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber || par.Type != ArgNumber || frequency.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() >= 7 {
|
|
if basis = argsList.Front().Next().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
if argsList.Len() == 8 {
|
|
if cm := argsList.Back().Value.(formulaArg).ToBool(); cm.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
frac1 := yearFrac(issue.Number, settlement.Number, int(basis.Number))
|
|
if frac1.Type != ArgNumber {
|
|
return frac1
|
|
}
|
|
return newNumberFormulaArg(par.Number * rate.Number * frac1.Number)
|
|
}
|
|
|
|
// ACCRINTM function returns the accrued interest in a security that pays
|
|
// interest at maturity. The syntax of the function is:
|
|
//
|
|
// ACCRINTM(issue,settlement,rate,[par],[basis])
|
|
func (fn *formulaFuncs) ACCRINTM(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ACCRINTM requires 4 or 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
issue, settlement := args.List[0], args.List[1]
|
|
if settlement.Number < issue.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
rate := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
par := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber || par.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if par.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(issue.Number, settlement.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg(frac.Number * rate.Number * par.Number)
|
|
}
|
|
|
|
// prepareAmorArgs checking and prepare arguments for the formula functions
|
|
// AMORDEGRC and AMORLINC.
|
|
func (fn *formulaFuncs) prepareAmorArgs(name string, argsList *list.List) formulaArg {
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires cost to be number argument", name))
|
|
}
|
|
if cost.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires cost >= 0", name))
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(argsList.Front().Next().Value.(formulaArg))
|
|
datePurchased := fn.DATEVALUE(args)
|
|
if datePurchased.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
args.Init()
|
|
args.PushBack(argsList.Front().Next().Next().Value.(formulaArg))
|
|
firstPeriod := fn.DATEVALUE(args)
|
|
if firstPeriod.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if firstPeriod.Number < datePurchased.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
salvage := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if salvage.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if salvage.Number < 0 || salvage.Number > cost.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
period := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if period.Type != ArgNumber || period.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
rate := argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber || rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 7 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{cost, datePurchased, firstPeriod, salvage, period, rate, basis})
|
|
}
|
|
|
|
// AMORDEGRC function is provided for users of the French accounting system.
|
|
// The function calculates the prorated linear depreciation of an asset for a
|
|
// specified accounting period. The syntax of the function is:
|
|
//
|
|
// AMORDEGRC(cost,date_purchased,first_period,salvage,period,rate,[basis])
|
|
func (fn *formulaFuncs) AMORDEGRC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 6 && argsList.Len() != 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AMORDEGRC requires 6 or 7 arguments")
|
|
}
|
|
args := fn.prepareAmorArgs("AMORDEGRC", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
cost, datePurchased, firstPeriod, salvage, period, rate, basis := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5], args.List[6]
|
|
if rate.Number >= 0.5 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "AMORDEGRC requires rate to be < 0.5")
|
|
}
|
|
assetsLife, amorCoeff := 1/rate.Number, 2.5
|
|
if assetsLife < 3 {
|
|
amorCoeff = 1
|
|
} else if assetsLife < 5 {
|
|
amorCoeff = 1.5
|
|
} else if assetsLife <= 6 {
|
|
amorCoeff = 2
|
|
}
|
|
rate.Number *= amorCoeff
|
|
frac := yearFrac(datePurchased.Number, firstPeriod.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
nRate := float64(int((frac.Number * cost.Number * rate.Number) + 0.5))
|
|
cost.Number -= nRate
|
|
rest := cost.Number - salvage.Number
|
|
for n := 0; n < int(period.Number); n++ {
|
|
nRate = float64(int((cost.Number * rate.Number) + 0.5))
|
|
rest -= nRate
|
|
if rest < 0 {
|
|
switch int(period.Number) - n {
|
|
case 0:
|
|
case 1:
|
|
return newNumberFormulaArg(float64(int((cost.Number * 0.5) + 0.5)))
|
|
default:
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
}
|
|
cost.Number -= nRate
|
|
}
|
|
return newNumberFormulaArg(nRate)
|
|
}
|
|
|
|
// AMORLINC function is provided for users of the French accounting system.
|
|
// The function calculates the prorated linear depreciation of an asset for a
|
|
// specified accounting period. The syntax of the function is:
|
|
//
|
|
// AMORLINC(cost,date_purchased,first_period,salvage,period,rate,[basis])
|
|
func (fn *formulaFuncs) AMORLINC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 6 && argsList.Len() != 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "AMORLINC requires 6 or 7 arguments")
|
|
}
|
|
args := fn.prepareAmorArgs("AMORLINC", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
cost, datePurchased, firstPeriod, salvage, period, rate, basis := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5], args.List[6]
|
|
frac := yearFrac(datePurchased.Number, firstPeriod.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
rate1 := frac.Number * cost.Number * rate.Number
|
|
if period.Number == 0 {
|
|
return newNumberFormulaArg(rate1)
|
|
}
|
|
rate2 := cost.Number * rate.Number
|
|
delta := cost.Number - salvage.Number
|
|
periods := int((delta - rate1) / rate2)
|
|
if int(period.Number) <= periods {
|
|
return newNumberFormulaArg(rate2)
|
|
} else if int(period.Number)-1 == periods {
|
|
return newNumberFormulaArg(delta - rate2*float64(periods) - math.Nextafter(rate1, rate1))
|
|
}
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
|
|
// prepareCouponArgs checking and prepare arguments for the formula functions
|
|
// COUPDAYBS, COUPDAYS, COUPDAYSNC, COUPPCD, COUPNUM and COUPNCD.
|
|
func (fn *formulaFuncs) prepareCouponArgs(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 && argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 3 or 4 arguments", name))
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
if settlement.Number >= maturity.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, fmt.Sprintf("%s requires maturity > settlement", name))
|
|
}
|
|
frequency := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if frequency.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 4 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{settlement, maturity, frequency, basis})
|
|
}
|
|
|
|
// is30BasisMethod determine if the financial day count basis rules is 30/360
|
|
// methods.
|
|
func is30BasisMethod(basis int) bool {
|
|
return basis == 0 || basis == 4
|
|
}
|
|
|
|
// getDaysInMonthRange return the day by given year, month range and day count
|
|
// basis.
|
|
func getDaysInMonthRange(fromMonth, toMonth int) int {
|
|
if fromMonth > toMonth {
|
|
return 0
|
|
}
|
|
return (toMonth - fromMonth + 1) * 30
|
|
}
|
|
|
|
// getDayOnBasis returns the day by given date and day count basis.
|
|
func getDayOnBasis(y, m, d, basis int) int {
|
|
if !is30BasisMethod(basis) {
|
|
return d
|
|
}
|
|
day := d
|
|
dim := getDaysInMonth(y, m)
|
|
if day > 30 || d >= dim || day >= dim {
|
|
day = 30
|
|
}
|
|
return day
|
|
}
|
|
|
|
// coupdays returns the number of days that base on date range and the day
|
|
// count basis to be used.
|
|
func coupdays(from, to time.Time, basis int) float64 {
|
|
days := 0
|
|
fromY, fromM, fromD := from.Date()
|
|
toY, toM, toD := to.Date()
|
|
fromDay, toDay := getDayOnBasis(fromY, int(fromM), fromD, basis), getDayOnBasis(toY, int(toM), toD, basis)
|
|
if !is30BasisMethod(basis) {
|
|
return (daysBetween(excelMinTime1900.Unix(), makeDate(toY, toM, toDay)) + 1) - (daysBetween(excelMinTime1900.Unix(), makeDate(fromY, fromM, fromDay)) + 1)
|
|
}
|
|
if basis == 0 {
|
|
if (int(fromM) == 2 || fromDay < 30) && toD == 31 {
|
|
toDay = 31
|
|
}
|
|
} else {
|
|
if int(fromM) == 2 && fromDay == 30 {
|
|
fromDay = getDaysInMonth(fromY, 2)
|
|
}
|
|
if int(toM) == 2 && toDay == 30 {
|
|
toDay = getDaysInMonth(toY, 2)
|
|
}
|
|
}
|
|
if fromY < toY || (fromY == toY && int(fromM) < int(toM)) {
|
|
days = 30 - fromDay + 1
|
|
fromD = 1
|
|
fromDay = 1
|
|
date := time.Date(fromY, fromM, fromD, 0, 0, 0, 0, time.UTC).AddDate(0, 1, 0)
|
|
if date.Year() < toY {
|
|
days += getDaysInMonthRange(int(date.Month()), 12)
|
|
date = date.AddDate(0, 13-int(date.Month()), 0)
|
|
}
|
|
days += getDaysInMonthRange(int(date.Month()), int(toM)-1)
|
|
}
|
|
if days += toDay - fromDay; days > 0 {
|
|
return float64(days)
|
|
}
|
|
return 0
|
|
}
|
|
|
|
// COUPDAYBS function calculates the number of days from the beginning of a
|
|
// coupon's period to the settlement date. The syntax of the function is:
|
|
//
|
|
// COUPDAYBS(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPDAYBS(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPDAYBS", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement := timeFromExcelTime(args.List[0].Number, false)
|
|
pcd := timeFromExcelTime(fn.COUPPCD(argsList).Number, false)
|
|
return newNumberFormulaArg(coupdays(pcd, settlement, int(args.List[3].Number)))
|
|
}
|
|
|
|
// COUPDAYS function calculates the number of days in a coupon period that
|
|
// contains the settlement date. The syntax of the function is:
|
|
//
|
|
// COUPDAYS(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPDAYS(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPDAYS", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
freq := args.List[2].Number
|
|
basis := int(args.List[3].Number)
|
|
if basis == 1 {
|
|
pcd := timeFromExcelTime(fn.COUPPCD(argsList).Number, false)
|
|
next := pcd.AddDate(0, 12/int(freq), 0)
|
|
return newNumberFormulaArg(coupdays(pcd, next, basis))
|
|
}
|
|
return newNumberFormulaArg(float64(getYearDays(0, basis)) / freq)
|
|
}
|
|
|
|
// COUPDAYSNC function calculates the number of days from the settlement date
|
|
// to the next coupon date. The syntax of the function is:
|
|
//
|
|
// COUPDAYSNC(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPDAYSNC(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPDAYSNC", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement := timeFromExcelTime(args.List[0].Number, false)
|
|
basis := int(args.List[3].Number)
|
|
ncd := timeFromExcelTime(fn.COUPNCD(argsList).Number, false)
|
|
return newNumberFormulaArg(coupdays(settlement, ncd, basis))
|
|
}
|
|
|
|
// coupons is an implementation of the formula functions COUPNCD and COUPPCD.
|
|
func (fn *formulaFuncs) coupons(name string, arg formulaArg) formulaArg {
|
|
settlement := timeFromExcelTime(arg.List[0].Number, false)
|
|
maturity := timeFromExcelTime(arg.List[1].Number, false)
|
|
maturityDays := (maturity.Year()-settlement.Year())*12 + (int(maturity.Month()) - int(settlement.Month()))
|
|
coupon := 12 / int(arg.List[2].Number)
|
|
mod := maturityDays % coupon
|
|
year := settlement.Year()
|
|
month := int(settlement.Month())
|
|
if mod == 0 && settlement.Day() >= maturity.Day() {
|
|
month += coupon
|
|
} else {
|
|
month += mod
|
|
}
|
|
if name != "COUPNCD" {
|
|
month -= coupon
|
|
}
|
|
if month > 11 {
|
|
year++
|
|
month -= 12
|
|
} else if month < 0 {
|
|
year--
|
|
month += 12
|
|
}
|
|
day, lastDay := maturity.Day(), time.Date(year, time.Month(month), 1, 0, 0, 0, 0, time.UTC)
|
|
days := getDaysInMonth(lastDay.Year(), int(lastDay.Month()))
|
|
if getDaysInMonth(maturity.Year(), int(maturity.Month())) == maturity.Day() {
|
|
day = days
|
|
} else if day > 27 && day > days {
|
|
day = days
|
|
}
|
|
return newNumberFormulaArg(daysBetween(excelMinTime1900.Unix(), makeDate(year, time.Month(month), day)) + 1)
|
|
}
|
|
|
|
// COUPNCD function calculates the number of coupons payable, between a
|
|
// security's settlement date and maturity date, rounded up to the nearest
|
|
// whole coupon. The syntax of the function is:
|
|
//
|
|
// COUPNCD(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPNCD(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPNCD", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
return fn.coupons("COUPNCD", args)
|
|
}
|
|
|
|
// COUPNUM function calculates the number of coupons payable, between a
|
|
// security's settlement date and maturity date, rounded up to the nearest
|
|
// whole coupon. The syntax of the function is:
|
|
//
|
|
// COUPNUM(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPNUM(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPNUM", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
frac := yearFrac(args.List[0].Number, args.List[1].Number, 0)
|
|
return newNumberFormulaArg(math.Ceil(frac.Number * args.List[2].Number))
|
|
}
|
|
|
|
// COUPPCD function returns the previous coupon date, before the settlement
|
|
// date for a security. The syntax of the function is:
|
|
//
|
|
// COUPPCD(settlement,maturity,frequency,[basis])
|
|
func (fn *formulaFuncs) COUPPCD(argsList *list.List) formulaArg {
|
|
args := fn.prepareCouponArgs("COUPPCD", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
return fn.coupons("COUPPCD", args)
|
|
}
|
|
|
|
// CUMIPMT function calculates the cumulative interest paid on a loan or
|
|
// investment, between two specified periods. The syntax of the function is:
|
|
//
|
|
// CUMIPMT(rate,nper,pv,start_period,end_period,type)
|
|
func (fn *formulaFuncs) CUMIPMT(argsList *list.List) formulaArg {
|
|
return fn.cumip("CUMIPMT", argsList)
|
|
}
|
|
|
|
// CUMPRINC function calculates the cumulative payment on the principal of a
|
|
// loan or investment, between two specified periods. The syntax of the
|
|
// function is:
|
|
//
|
|
// CUMPRINC(rate,nper,pv,start_period,end_period,type)
|
|
func (fn *formulaFuncs) CUMPRINC(argsList *list.List) formulaArg {
|
|
return fn.cumip("CUMPRINC", argsList)
|
|
}
|
|
|
|
// cumip is an implementation of the formula functions CUMIPMT and CUMPRINC.
|
|
func (fn *formulaFuncs) cumip(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 6 arguments", name))
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
nper := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pv := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
start := argsList.Back().Prev().Prev().Value.(formulaArg).ToNumber()
|
|
if start.Type != ArgNumber {
|
|
return start
|
|
}
|
|
end := argsList.Back().Prev().Value.(formulaArg).ToNumber()
|
|
if end.Type != ArgNumber {
|
|
return end
|
|
}
|
|
typ := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if typ.Type != ArgNumber {
|
|
return typ
|
|
}
|
|
if typ.Number != 0 && typ.Number != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if start.Number < 1 || start.Number > end.Number {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
num := 0.0
|
|
for per := start.Number; per <= end.Number; per++ {
|
|
args := list.New().Init()
|
|
args.PushBack(rate)
|
|
args.PushBack(newNumberFormulaArg(per))
|
|
args.PushBack(nper)
|
|
args.PushBack(pv)
|
|
args.PushBack(newNumberFormulaArg(0))
|
|
args.PushBack(typ)
|
|
if name == "CUMIPMT" {
|
|
num += fn.IPMT(args).Number
|
|
continue
|
|
}
|
|
num += fn.PPMT(args).Number
|
|
}
|
|
return newNumberFormulaArg(num)
|
|
}
|
|
|
|
// calcDbArgsCompare implements common arguments' comparison for DB and DDB.
|
|
func calcDbArgsCompare(cost, salvage, life, period formulaArg) bool {
|
|
return (cost.Number <= 0) || ((salvage.Number / cost.Number) < 0) || (life.Number <= 0) || (period.Number < 1)
|
|
}
|
|
|
|
// DB function calculates the depreciation of an asset, using the Fixed
|
|
// Declining Balance Method, for each period of the asset's lifetime. The
|
|
// syntax of the function is:
|
|
//
|
|
// DB(cost,salvage,life,period,[month])
|
|
func (fn *formulaFuncs) DB(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DB requires at least 4 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DB allows at most 5 arguments")
|
|
}
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber {
|
|
return cost
|
|
}
|
|
salvage := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if salvage.Type != ArgNumber {
|
|
return salvage
|
|
}
|
|
life := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if life.Type != ArgNumber {
|
|
return life
|
|
}
|
|
period := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if period.Type != ArgNumber {
|
|
return period
|
|
}
|
|
month := newNumberFormulaArg(12)
|
|
if argsList.Len() == 5 {
|
|
if month = argsList.Back().Value.(formulaArg).ToNumber(); month.Type != ArgNumber {
|
|
return month
|
|
}
|
|
}
|
|
if cost.Number == 0 {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
if calcDbArgsCompare(cost, salvage, life, period) || (month.Number < 1) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
dr := 1 - math.Pow(salvage.Number/cost.Number, 1/life.Number)
|
|
dr = math.Round(dr*1000) / 1000
|
|
pd, depreciation := 0.0, 0.0
|
|
for per := 1; per <= int(period.Number); per++ {
|
|
if per == 1 {
|
|
depreciation = cost.Number * dr * month.Number / 12
|
|
} else if per == int(life.Number+1) {
|
|
depreciation = (cost.Number - pd) * dr * (12 - month.Number) / 12
|
|
} else {
|
|
depreciation = (cost.Number - pd) * dr
|
|
}
|
|
pd += depreciation
|
|
}
|
|
return newNumberFormulaArg(depreciation)
|
|
}
|
|
|
|
// DDB function calculates the depreciation of an asset, using the Double
|
|
// Declining Balance Method, or another specified depreciation rate. The
|
|
// syntax of the function is:
|
|
//
|
|
// DDB(cost,salvage,life,period,[factor])
|
|
func (fn *formulaFuncs) DDB(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DDB requires at least 4 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DDB allows at most 5 arguments")
|
|
}
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber {
|
|
return cost
|
|
}
|
|
salvage := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if salvage.Type != ArgNumber {
|
|
return salvage
|
|
}
|
|
life := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if life.Type != ArgNumber {
|
|
return life
|
|
}
|
|
period := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if period.Type != ArgNumber {
|
|
return period
|
|
}
|
|
factor := newNumberFormulaArg(2)
|
|
if argsList.Len() == 5 {
|
|
if factor = argsList.Back().Value.(formulaArg).ToNumber(); factor.Type != ArgNumber {
|
|
return factor
|
|
}
|
|
}
|
|
if cost.Number == 0 {
|
|
return newNumberFormulaArg(0)
|
|
}
|
|
if calcDbArgsCompare(cost, salvage, life, period) || (factor.Number <= 0.0) || (period.Number > life.Number) {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
pd, depreciation := 0.0, 0.0
|
|
for per := 1; per <= int(period.Number); per++ {
|
|
depreciation = math.Min((cost.Number-pd)*(factor.Number/life.Number), cost.Number-salvage.Number-pd)
|
|
pd += depreciation
|
|
}
|
|
return newNumberFormulaArg(depreciation)
|
|
}
|
|
|
|
// prepareDataValueArgs convert first N arguments to data value for the
|
|
// formula functions.
|
|
func (fn *formulaFuncs) prepareDataValueArgs(n int, argsList *list.List) formulaArg {
|
|
l := list.New()
|
|
var dataValues []formulaArg
|
|
getDateValue := func(arg formulaArg, l *list.List) formulaArg {
|
|
switch arg.Type {
|
|
case ArgNumber:
|
|
break
|
|
case ArgString:
|
|
num := arg.ToNumber()
|
|
if num.Type == ArgNumber {
|
|
arg = num
|
|
break
|
|
}
|
|
l.Init()
|
|
l.PushBack(arg)
|
|
arg = fn.DATEVALUE(l)
|
|
if arg.Type == ArgError {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
default:
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
return arg
|
|
}
|
|
for i, arg := 0, argsList.Front(); i < n; arg = arg.Next() {
|
|
dataValue := getDateValue(arg.Value.(formulaArg), l)
|
|
if dataValue.Type != ArgNumber {
|
|
return dataValue
|
|
}
|
|
dataValues = append(dataValues, dataValue)
|
|
i++
|
|
}
|
|
return newListFormulaArg(dataValues)
|
|
}
|
|
|
|
// DISC function calculates the Discount Rate for a security. The syntax of
|
|
// the function is:
|
|
//
|
|
// DISC(settlement,maturity,pr,redemption,[basis])
|
|
func (fn *formulaFuncs) DISC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DISC requires 4 or 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
if maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "DISC requires maturity > settlement")
|
|
}
|
|
pr := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pr.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if pr.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "DISC requires pr > 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "DISC requires redemption > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg((redemption.Number - pr.Number) / redemption.Number / frac.Number)
|
|
}
|
|
|
|
// DOLLARDE function converts a dollar value in fractional notation, into a
|
|
// dollar value expressed as a decimal. The syntax of the function is:
|
|
//
|
|
// DOLLARDE(fractional_dollar,fraction)
|
|
func (fn *formulaFuncs) DOLLARDE(argsList *list.List) formulaArg {
|
|
return fn.dollar("DOLLARDE", argsList)
|
|
}
|
|
|
|
// DOLLARFR function converts a dollar value in decimal notation, into a
|
|
// dollar value that is expressed in fractional notation. The syntax of the
|
|
// function is:
|
|
//
|
|
// DOLLARFR(decimal_dollar,fraction)
|
|
func (fn *formulaFuncs) DOLLARFR(argsList *list.List) formulaArg {
|
|
return fn.dollar("DOLLARFR", argsList)
|
|
}
|
|
|
|
// dollar is an implementation of the formula functions DOLLARDE and DOLLARFR.
|
|
func (fn *formulaFuncs) dollar(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 2 arguments", name))
|
|
}
|
|
dollar := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if dollar.Type != ArgNumber {
|
|
return dollar
|
|
}
|
|
frac := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
if frac.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if frac.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
cents := math.Mod(dollar.Number, 1)
|
|
if name == "DOLLARDE" {
|
|
cents /= frac.Number
|
|
cents *= math.Pow(10, math.Ceil(math.Log10(frac.Number)))
|
|
} else {
|
|
cents *= frac.Number
|
|
cents *= math.Pow(10, -math.Ceil(math.Log10(frac.Number)))
|
|
}
|
|
return newNumberFormulaArg(math.Floor(dollar.Number) + cents)
|
|
}
|
|
|
|
// prepareDurationArgs checking and prepare arguments for the formula
|
|
// functions DURATION and MDURATION.
|
|
func (fn *formulaFuncs) prepareDurationArgs(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 5 && argsList.Len() != 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 5 or 6 arguments", name))
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
if settlement.Number >= maturity.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, fmt.Sprintf("%s requires maturity > settlement", name))
|
|
}
|
|
coupon := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if coupon.Type != ArgNumber {
|
|
return coupon
|
|
}
|
|
if coupon.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, fmt.Sprintf("%s requires coupon >= 0", name))
|
|
}
|
|
yld := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if yld.Type != ArgNumber {
|
|
return yld
|
|
}
|
|
if yld.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, fmt.Sprintf("%s requires yld >= 0", name))
|
|
}
|
|
frequency := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if frequency.Type != ArgNumber {
|
|
return frequency
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 6 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{settlement, maturity, coupon, yld, frequency, basis})
|
|
}
|
|
|
|
// duration is an implementation of the formula function DURATION.
|
|
func (fn *formulaFuncs) duration(settlement, maturity, coupon, yld, frequency, basis formulaArg) formulaArg {
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
argumments := list.New().Init()
|
|
argumments.PushBack(settlement)
|
|
argumments.PushBack(maturity)
|
|
argumments.PushBack(frequency)
|
|
argumments.PushBack(basis)
|
|
coups := fn.COUPNUM(argumments)
|
|
duration := 0.0
|
|
p := 0.0
|
|
coupon.Number *= 100 / frequency.Number
|
|
yld.Number /= frequency.Number
|
|
yld.Number++
|
|
diff := frac.Number*frequency.Number - coups.Number
|
|
for t := 1.0; t < coups.Number; t++ {
|
|
tDiff := t + diff
|
|
add := coupon.Number / math.Pow(yld.Number, tDiff)
|
|
p += add
|
|
duration += tDiff * add
|
|
}
|
|
add := (coupon.Number + 100) / math.Pow(yld.Number, coups.Number+diff)
|
|
p += add
|
|
duration += (coups.Number + diff) * add
|
|
duration /= p
|
|
duration /= frequency.Number
|
|
return newNumberFormulaArg(duration)
|
|
}
|
|
|
|
// DURATION function calculates the Duration (specifically, the Macaulay
|
|
// Duration) of a security that pays periodic interest, assuming a par value
|
|
// of $100. The syntax of the function is:
|
|
//
|
|
// DURATION(settlement,maturity,coupon,yld,frequency,[basis])
|
|
func (fn *formulaFuncs) DURATION(argsList *list.List) formulaArg {
|
|
args := fn.prepareDurationArgs("DURATION", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
return fn.duration(args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5])
|
|
}
|
|
|
|
// EFFECT function returns the effective annual interest rate for a given
|
|
// nominal interest rate and number of compounding periods per year. The
|
|
// syntax of the function is:
|
|
//
|
|
// EFFECT(nominal_rate,npery)
|
|
func (fn *formulaFuncs) EFFECT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EFFECT requires 2 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
npery := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if npery.Type != ArgNumber {
|
|
return npery
|
|
}
|
|
if rate.Number <= 0 || npery.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(math.Pow(1+rate.Number/npery.Number, npery.Number) - 1)
|
|
}
|
|
|
|
// EUROCONVERT function convert a number to euro or from euro to a
|
|
// participating currency. You can also use it to convert a number from one
|
|
// participating currency to another by using the euro as an intermediary
|
|
// (triangulation). The syntax of the function is:
|
|
//
|
|
// EUROCONVERT(number,sourcecurrency,targetcurrency[,fullprecision,triangulationprecision])
|
|
func (fn *formulaFuncs) EUROCONVERT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EUROCONVERT requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "EUROCONVERT allows at most 5 arguments")
|
|
}
|
|
number := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if number.Type != ArgNumber {
|
|
return number
|
|
}
|
|
sourceCurrency := argsList.Front().Next().Value.(formulaArg).Value()
|
|
targetCurrency := argsList.Front().Next().Next().Value.(formulaArg).Value()
|
|
fullPrec, triangulationPrec := newBoolFormulaArg(false), newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if fullPrec = argsList.Front().Next().Next().Next().Value.(formulaArg).ToBool(); fullPrec.Type != ArgNumber {
|
|
return fullPrec
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if triangulationPrec = argsList.Back().Value.(formulaArg).ToNumber(); triangulationPrec.Type != ArgNumber {
|
|
return triangulationPrec
|
|
}
|
|
}
|
|
convertTable := map[string][]float64{
|
|
"EUR": {1.0, 2},
|
|
"ATS": {13.7603, 2},
|
|
"BEF": {40.3399, 0},
|
|
"DEM": {1.95583, 2},
|
|
"ESP": {166.386, 0},
|
|
"FIM": {5.94573, 2},
|
|
"FRF": {6.55957, 2},
|
|
"IEP": {0.787564, 2},
|
|
"ITL": {1936.27, 0},
|
|
"LUF": {40.3399, 0},
|
|
"NLG": {2.20371, 2},
|
|
"PTE": {200.482, 2},
|
|
"GRD": {340.750, 2},
|
|
"SIT": {239.640, 2},
|
|
"MTL": {0.429300, 2},
|
|
"CYP": {0.585274, 2},
|
|
"SKK": {30.1260, 2},
|
|
"EEK": {15.6466, 2},
|
|
"LVL": {0.702804, 2},
|
|
"LTL": {3.45280, 2},
|
|
}
|
|
source, ok := convertTable[sourceCurrency]
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
target, ok := convertTable[targetCurrency]
|
|
if !ok {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if sourceCurrency == targetCurrency {
|
|
return number
|
|
}
|
|
var res float64
|
|
if sourceCurrency == "EUR" {
|
|
res = number.Number * target[0]
|
|
} else {
|
|
intermediate := number.Number / source[0]
|
|
if triangulationPrec.Number != 0 {
|
|
ratio := math.Pow(10, triangulationPrec.Number)
|
|
intermediate = math.Round(intermediate*ratio) / ratio
|
|
}
|
|
res = intermediate * target[0]
|
|
}
|
|
if fullPrec.Number != 1 {
|
|
ratio := math.Pow(10, target[1])
|
|
res = math.Round(res*ratio) / ratio
|
|
}
|
|
return newNumberFormulaArg(res)
|
|
}
|
|
|
|
// FV function calculates the Future Value of an investment with periodic
|
|
// constant payments and a constant interest rate. The syntax of the function
|
|
// is:
|
|
//
|
|
// FV(rate,nper,[pmt],[pv],[type])
|
|
func (fn *formulaFuncs) FV(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FV requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FV allows at most 5 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
nper := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pmt := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pmt.Type != ArgNumber {
|
|
return pmt
|
|
}
|
|
pv, typ := newNumberFormulaArg(0), newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if pv = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if typ = argsList.Back().Value.(formulaArg).ToNumber(); typ.Type != ArgNumber {
|
|
return typ
|
|
}
|
|
}
|
|
if typ.Number != 0 && typ.Number != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if rate.Number != 0 {
|
|
return newNumberFormulaArg(-pv.Number*math.Pow(1+rate.Number, nper.Number) - pmt.Number*(1+rate.Number*typ.Number)*(math.Pow(1+rate.Number, nper.Number)-1)/rate.Number)
|
|
}
|
|
return newNumberFormulaArg(-pv.Number - pmt.Number*nper.Number)
|
|
}
|
|
|
|
// FVSCHEDULE function calculates the Future Value of an investment with a
|
|
// variable interest rate. The syntax of the function is:
|
|
//
|
|
// FVSCHEDULE(principal,schedule)
|
|
func (fn *formulaFuncs) FVSCHEDULE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "FVSCHEDULE requires 2 arguments")
|
|
}
|
|
pri := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if pri.Type != ArgNumber {
|
|
return pri
|
|
}
|
|
principal := pri.Number
|
|
for _, arg := range argsList.Back().Value.(formulaArg).ToList() {
|
|
if arg.Value() == "" {
|
|
continue
|
|
}
|
|
rate := arg.ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
principal *= 1 + rate.Number
|
|
}
|
|
return newNumberFormulaArg(principal)
|
|
}
|
|
|
|
// INTRATE function calculates the interest rate for a fully invested
|
|
// security. The syntax of the function is:
|
|
//
|
|
// INTRATE(settlement,maturity,investment,redemption,[basis])
|
|
func (fn *formulaFuncs) INTRATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "INTRATE requires 4 or 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
if maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "INTRATE requires maturity > settlement")
|
|
}
|
|
investment := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if investment.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if investment.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "INTRATE requires investment > 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "INTRATE requires redemption > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg((redemption.Number - investment.Number) / investment.Number / frac.Number)
|
|
}
|
|
|
|
// IPMT function calculates the interest payment, during a specific period of a
|
|
// loan or investment that is paid in constant periodic payments, with a
|
|
// constant interest rate. The syntax of the function is:
|
|
//
|
|
// IPMT(rate,per,nper,pv,[fv],[type])
|
|
func (fn *formulaFuncs) IPMT(argsList *list.List) formulaArg {
|
|
return fn.ipmt("IPMT", argsList)
|
|
}
|
|
|
|
// calcIpmt is part of the implementation ipmt.
|
|
func calcIpmt(name string, typ, per, pmt, pv, rate formulaArg) formulaArg {
|
|
capital, interest, principal := pv.Number, 0.0, 0.0
|
|
for i := 1; i <= int(per.Number); i++ {
|
|
if typ.Number != 0 && i == 1 {
|
|
interest = 0
|
|
} else {
|
|
interest = -capital * rate.Number
|
|
}
|
|
principal = pmt.Number - interest
|
|
capital += principal
|
|
}
|
|
if name == "IPMT" {
|
|
return newNumberFormulaArg(interest)
|
|
}
|
|
return newNumberFormulaArg(principal)
|
|
}
|
|
|
|
// ipmt is an implementation of the formula functions IPMT and PPMT.
|
|
func (fn *formulaFuncs) ipmt(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 4 arguments", name))
|
|
}
|
|
if argsList.Len() > 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 6 arguments", name))
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
per := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if per.Type != ArgNumber {
|
|
return per
|
|
}
|
|
nper := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pv := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
fv, typ := newNumberFormulaArg(0), newNumberFormulaArg(0)
|
|
if argsList.Len() >= 5 {
|
|
if fv = argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
}
|
|
if argsList.Len() == 6 {
|
|
if typ = argsList.Back().Value.(formulaArg).ToNumber(); typ.Type != ArgNumber {
|
|
return typ
|
|
}
|
|
}
|
|
if typ.Number != 0 && typ.Number != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if per.Number <= 0 || per.Number > nper.Number {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
args := list.New().Init()
|
|
args.PushBack(rate)
|
|
args.PushBack(nper)
|
|
args.PushBack(pv)
|
|
args.PushBack(fv)
|
|
args.PushBack(typ)
|
|
pmt := fn.PMT(args)
|
|
return calcIpmt(name, typ, per, pmt, pv, rate)
|
|
}
|
|
|
|
// IRR function returns the Internal Rate of Return for a supplied series of
|
|
// periodic cash flows (i.e. an initial investment value and a series of net
|
|
// income values). The syntax of the function is:
|
|
//
|
|
// IRR(values,[guess])
|
|
func (fn *formulaFuncs) IRR(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IRR requires at least 1 argument")
|
|
}
|
|
if argsList.Len() > 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "IRR allows at most 2 arguments")
|
|
}
|
|
values, guess := argsList.Front().Value.(formulaArg).ToList(), newNumberFormulaArg(0.1)
|
|
if argsList.Len() > 1 {
|
|
if guess = argsList.Back().Value.(formulaArg).ToNumber(); guess.Type != ArgNumber {
|
|
return guess
|
|
}
|
|
}
|
|
x1, x2 := newNumberFormulaArg(0), guess
|
|
args := list.New().Init()
|
|
args.PushBack(x1)
|
|
for _, v := range values {
|
|
args.PushBack(v)
|
|
}
|
|
f1 := fn.NPV(args)
|
|
args.Front().Value = x2
|
|
f2 := fn.NPV(args)
|
|
for i := 0; i < maxFinancialIterations; i++ {
|
|
if f1.Number*f2.Number < 0 {
|
|
break
|
|
}
|
|
if math.Abs(f1.Number) < math.Abs(f2.Number) {
|
|
x1.Number += 1.6 * (x1.Number - x2.Number)
|
|
args.Front().Value = x1
|
|
f1 = fn.NPV(args)
|
|
continue
|
|
}
|
|
x2.Number += 1.6 * (x2.Number - x1.Number)
|
|
args.Front().Value = x2
|
|
f2 = fn.NPV(args)
|
|
}
|
|
if f1.Number*f2.Number > 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
args.Front().Value = x1
|
|
f := fn.NPV(args)
|
|
var rtb, dx, xMid, fMid float64
|
|
if f.Number < 0 {
|
|
rtb = x1.Number
|
|
dx = x2.Number - x1.Number
|
|
} else {
|
|
rtb = x2.Number
|
|
dx = x1.Number - x2.Number
|
|
}
|
|
for i := 0; i < maxFinancialIterations; i++ {
|
|
dx *= 0.5
|
|
xMid = rtb + dx
|
|
args.Front().Value = newNumberFormulaArg(xMid)
|
|
fMid = fn.NPV(args).Number
|
|
if fMid <= 0 {
|
|
rtb = xMid
|
|
}
|
|
if math.Abs(fMid) < financialPrecision || math.Abs(dx) < financialPrecision {
|
|
break
|
|
}
|
|
}
|
|
return newNumberFormulaArg(xMid)
|
|
}
|
|
|
|
// ISPMT function calculates the interest paid during a specific period of a
|
|
// loan or investment. The syntax of the function is:
|
|
//
|
|
// ISPMT(rate,per,nper,pv)
|
|
func (fn *formulaFuncs) ISPMT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ISPMT requires 4 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
per := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if per.Type != ArgNumber {
|
|
return per
|
|
}
|
|
nper := argsList.Back().Prev().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pv := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
pr, payment, num := pv.Number, pv.Number/nper.Number, 0.0
|
|
for i := 0; i <= int(per.Number); i++ {
|
|
num = rate.Number * pr * -1
|
|
pr -= payment
|
|
if i == int(nper.Number) {
|
|
num = 0
|
|
}
|
|
}
|
|
return newNumberFormulaArg(num)
|
|
}
|
|
|
|
// MDURATION function calculates the Modified Macaulay Duration of a security
|
|
// that pays periodic interest, assuming a par value of $100. The syntax of
|
|
// the function is:
|
|
//
|
|
// MDURATION(settlement,maturity,coupon,yld,frequency,[basis])
|
|
func (fn *formulaFuncs) MDURATION(argsList *list.List) formulaArg {
|
|
args := fn.prepareDurationArgs("MDURATION", argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
duration := fn.duration(args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5])
|
|
if duration.Type != ArgNumber {
|
|
return duration
|
|
}
|
|
return newNumberFormulaArg(duration.Number / (1 + args.List[3].Number/args.List[4].Number))
|
|
}
|
|
|
|
// MIRR function returns the Modified Internal Rate of Return for a supplied
|
|
// series of periodic cash flows (i.e. a set of values, which includes an
|
|
// initial investment value and a series of net income values). The syntax of
|
|
// the function is:
|
|
//
|
|
// MIRR(values,finance_rate,reinvest_rate)
|
|
func (fn *formulaFuncs) MIRR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "MIRR requires 3 arguments")
|
|
}
|
|
values := argsList.Front().Value.(formulaArg).ToList()
|
|
financeRate := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if financeRate.Type != ArgNumber {
|
|
return financeRate
|
|
}
|
|
reinvestRate := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if reinvestRate.Type != ArgNumber {
|
|
return reinvestRate
|
|
}
|
|
n, fr, rr, npvPos, npvNeg := len(values), 1+financeRate.Number, 1+reinvestRate.Number, 0.0, 0.0
|
|
for i, v := range values {
|
|
val := v.ToNumber()
|
|
if val.Number >= 0 {
|
|
npvPos += val.Number / math.Pow(rr, float64(i))
|
|
continue
|
|
}
|
|
npvNeg += val.Number / math.Pow(fr, float64(i))
|
|
}
|
|
if npvNeg == 0 || npvPos == 0 || reinvestRate.Number <= -1 {
|
|
return newErrorFormulaArg(formulaErrorDIV, formulaErrorDIV)
|
|
}
|
|
return newNumberFormulaArg(math.Pow(-npvPos*math.Pow(rr, float64(n))/(npvNeg*rr), 1/(float64(n)-1)) - 1)
|
|
}
|
|
|
|
// NOMINAL function returns the nominal interest rate for a given effective
|
|
// interest rate and number of compounding periods per year. The syntax of
|
|
// the function is:
|
|
//
|
|
// NOMINAL(effect_rate,npery)
|
|
func (fn *formulaFuncs) NOMINAL(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NOMINAL requires 2 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
npery := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if npery.Type != ArgNumber {
|
|
return npery
|
|
}
|
|
if rate.Number <= 0 || npery.Number < 1 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(npery.Number * (math.Pow(rate.Number+1, 1/npery.Number) - 1))
|
|
}
|
|
|
|
// NPER function calculates the number of periods required to pay off a loan,
|
|
// for a constant periodic payment and a constant interest rate. The syntax
|
|
// of the function is:
|
|
//
|
|
// NPER(rate,pmt,pv,[fv],[type])
|
|
func (fn *formulaFuncs) NPER(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NPER requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NPER allows at most 5 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
pmt := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if pmt.Type != ArgNumber {
|
|
return pmt
|
|
}
|
|
pv := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
fv, typ := newNumberFormulaArg(0), newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if fv = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if typ = argsList.Back().Value.(formulaArg).ToNumber(); typ.Type != ArgNumber {
|
|
return typ
|
|
}
|
|
}
|
|
if typ.Number != 0 && typ.Number != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if pmt.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if rate.Number != 0 {
|
|
p := math.Log((pmt.Number*(1+rate.Number*typ.Number)/rate.Number-fv.Number)/(pv.Number+pmt.Number*(1+rate.Number*typ.Number)/rate.Number)) / math.Log(1+rate.Number)
|
|
return newNumberFormulaArg(p)
|
|
}
|
|
return newNumberFormulaArg((-pv.Number - fv.Number) / pmt.Number)
|
|
}
|
|
|
|
// NPV function calculates the Net Present Value of an investment, based on a
|
|
// supplied discount rate, and a series of future payments and income. The
|
|
// syntax of the function is:
|
|
//
|
|
// NPV(rate,value1,[value2],[value3],...)
|
|
func (fn *formulaFuncs) NPV(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "NPV requires at least 2 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
val, i := 0.0, 1
|
|
for arg := argsList.Front().Next(); arg != nil; arg = arg.Next() {
|
|
num := arg.Value.(formulaArg).ToNumber()
|
|
if num.Type != ArgNumber {
|
|
continue
|
|
}
|
|
val += num.Number / math.Pow(1+rate.Number, float64(i))
|
|
i++
|
|
}
|
|
return newNumberFormulaArg(val)
|
|
}
|
|
|
|
// aggrBetween is a part of implementation of the formula function ODDFPRICE.
|
|
func aggrBetween(startPeriod, endPeriod float64, initialValue []float64, f func(acc []float64, index float64) []float64) []float64 {
|
|
var s []float64
|
|
if startPeriod <= endPeriod {
|
|
for i := startPeriod; i <= endPeriod; i++ {
|
|
s = append(s, i)
|
|
}
|
|
} else {
|
|
for i := startPeriod; i >= endPeriod; i-- {
|
|
s = append(s, i)
|
|
}
|
|
}
|
|
return fold(f, initialValue, s)
|
|
}
|
|
|
|
// fold is a part of implementation of the formula function ODDFPRICE.
|
|
func fold(f func(acc []float64, index float64) []float64, state []float64, source []float64) []float64 {
|
|
length, value := len(source), state
|
|
for index := 0; length > index; index++ {
|
|
value = f(value, source[index])
|
|
}
|
|
return value
|
|
}
|
|
|
|
// changeMonth is a part of implementation of the formula function ODDFPRICE.
|
|
func changeMonth(date time.Time, numMonths float64, returnLastMonth bool) time.Time {
|
|
offsetDay := 0
|
|
if returnLastMonth && date.Day() == getDaysInMonth(date.Year(), int(date.Month())) {
|
|
offsetDay--
|
|
}
|
|
newDate := date.AddDate(0, int(numMonths), offsetDay)
|
|
if returnLastMonth {
|
|
lastDay := getDaysInMonth(newDate.Year(), int(newDate.Month()))
|
|
return timeFromExcelTime(daysBetween(excelMinTime1900.Unix(), makeDate(newDate.Year(), newDate.Month(), lastDay))+1, false)
|
|
}
|
|
return newDate
|
|
}
|
|
|
|
// datesAggregate is a part of implementation of the formula function
|
|
// ODDFPRICE.
|
|
func datesAggregate(startDate, endDate time.Time, numMonths float64, f func(pcd, ncd time.Time) float64, acc float64, returnLastMonth bool) (time.Time, time.Time, float64) {
|
|
frontDate, trailingDate := startDate, endDate
|
|
s1 := frontDate.After(endDate) || frontDate.Equal(endDate)
|
|
s2 := endDate.After(frontDate) || endDate.Equal(frontDate)
|
|
stop := s2
|
|
if numMonths > 0 {
|
|
stop = s1
|
|
}
|
|
for !stop {
|
|
trailingDate = frontDate
|
|
frontDate = changeMonth(frontDate, numMonths, returnLastMonth)
|
|
fn := f(frontDate, trailingDate)
|
|
acc += fn
|
|
s1 = frontDate.After(endDate) || frontDate.Equal(endDate)
|
|
s2 = endDate.After(frontDate) || endDate.Equal(frontDate)
|
|
stop = s2
|
|
if numMonths > 0 {
|
|
stop = s1
|
|
}
|
|
}
|
|
return frontDate, trailingDate, acc
|
|
}
|
|
|
|
// coupNumber is a part of implementation of the formula function ODDFPRICE.
|
|
func coupNumber(maturity, settlement, numMonths float64) float64 {
|
|
maturityTime, settlementTime := timeFromExcelTime(maturity, false), timeFromExcelTime(settlement, false)
|
|
my, mm, md := maturityTime.Year(), maturityTime.Month(), maturityTime.Day()
|
|
sy, sm, sd := settlementTime.Year(), settlementTime.Month(), settlementTime.Day()
|
|
couponsTemp, endOfMonthTemp := 0.0, getDaysInMonth(my, int(mm)) == md
|
|
endOfMonth := endOfMonthTemp
|
|
if !endOfMonthTemp && mm != 2 && md > 28 && md < getDaysInMonth(my, int(mm)) {
|
|
endOfMonth = getDaysInMonth(sy, int(sm)) == sd
|
|
}
|
|
startDate := changeMonth(settlementTime, 0, endOfMonth)
|
|
coupons := couponsTemp
|
|
if startDate.After(settlementTime) {
|
|
coupons++
|
|
}
|
|
date := changeMonth(startDate, numMonths, endOfMonth)
|
|
f := func(pcd, ncd time.Time) float64 {
|
|
return 1
|
|
}
|
|
_, _, result := datesAggregate(date, maturityTime, numMonths, f, coupons, endOfMonth)
|
|
return result
|
|
}
|
|
|
|
// prepareOddfpriceArgs checking and prepare arguments for the formula
|
|
// function ODDFPRICE.
|
|
func (fn *formulaFuncs) prepareOddfpriceArgs(argsList *list.List) formulaArg {
|
|
dateValues := fn.prepareDataValueArgs(4, argsList)
|
|
if dateValues.Type != ArgList {
|
|
return dateValues
|
|
}
|
|
settlement, maturity, issue, firstCoupon := dateValues.List[0], dateValues.List[1], dateValues.List[2], dateValues.List[3]
|
|
if issue.Number >= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires settlement > issue")
|
|
}
|
|
if settlement.Number >= firstCoupon.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires first_coupon > settlement")
|
|
}
|
|
if firstCoupon.Number >= maturity.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires maturity > first_coupon")
|
|
}
|
|
rate := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires rate >= 0")
|
|
}
|
|
yld := argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if yld.Type != ArgNumber {
|
|
return yld
|
|
}
|
|
if yld.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires yld >= 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return redemption
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "ODDFPRICE requires redemption > 0")
|
|
}
|
|
frequency := argsList.Front().Next().Next().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if frequency.Type != ArgNumber {
|
|
return frequency
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 9 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{settlement, maturity, issue, firstCoupon, rate, yld, redemption, frequency, basis})
|
|
}
|
|
|
|
// ODDFPRICE function calculates the price per $100 face value of a security
|
|
// with an odd (short or long) first period. The syntax of the function is:
|
|
//
|
|
// ODDFPRICE(settlement,maturity,issue,first_coupon,rate,yld,redemption,frequency,[basis])
|
|
func (fn *formulaFuncs) ODDFPRICE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 8 && argsList.Len() != 9 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "ODDFPRICE requires 8 or 9 arguments")
|
|
}
|
|
args := fn.prepareOddfpriceArgs(argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity, issue, firstCoupon, rate, yld, redemption, frequency, basisArg := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5], args.List[6], args.List[7], args.List[8]
|
|
if basisArg.Number < 0 || basisArg.Number > 4 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "invalid basis")
|
|
}
|
|
issueTime := timeFromExcelTime(issue.Number, false)
|
|
settlementTime := timeFromExcelTime(settlement.Number, false)
|
|
maturityTime := timeFromExcelTime(maturity.Number, false)
|
|
firstCouponTime := timeFromExcelTime(firstCoupon.Number, false)
|
|
basis := int(basisArg.Number)
|
|
monthDays := getDaysInMonth(maturityTime.Year(), int(maturityTime.Month()))
|
|
returnLastMonth := monthDays == maturityTime.Day()
|
|
numMonths := 12 / frequency.Number
|
|
numMonthsNeg := -numMonths
|
|
mat := changeMonth(maturityTime, numMonthsNeg, returnLastMonth)
|
|
pcd, _, _ := datesAggregate(mat, firstCouponTime, numMonthsNeg, func(d1, d2 time.Time) float64 {
|
|
return 0
|
|
}, 0, returnLastMonth)
|
|
if !pcd.Equal(firstCouponTime) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
fnArgs := list.New().Init()
|
|
fnArgs.PushBack(settlement)
|
|
fnArgs.PushBack(maturity)
|
|
fnArgs.PushBack(frequency)
|
|
fnArgs.PushBack(basisArg)
|
|
e := fn.COUPDAYS(fnArgs)
|
|
n := fn.COUPNUM(fnArgs)
|
|
m := frequency.Number
|
|
dfc := coupdays(issueTime, firstCouponTime, basis)
|
|
if dfc < e.Number {
|
|
dsc := coupdays(settlementTime, firstCouponTime, basis)
|
|
a := coupdays(issueTime, settlementTime, basis)
|
|
x := yld.Number/m + 1
|
|
y := dsc / e.Number
|
|
p1 := x
|
|
p3 := math.Pow(p1, n.Number-1+y)
|
|
term1 := redemption.Number / p3
|
|
term2 := 100 * rate.Number / m * dfc / e.Number / math.Pow(p1, y)
|
|
f := func(acc []float64, index float64) []float64 {
|
|
return []float64{acc[0] + 100*rate.Number/m/math.Pow(p1, index-1+y)}
|
|
}
|
|
term3 := aggrBetween(2, math.Floor(n.Number), []float64{0}, f)
|
|
p2 := rate.Number / m
|
|
term4 := a / e.Number * p2 * 100
|
|
return newNumberFormulaArg(term1 + term2 + term3[0] - term4)
|
|
}
|
|
fnArgs.Init()
|
|
fnArgs.PushBack(issue)
|
|
fnArgs.PushBack(firstCoupon)
|
|
fnArgs.PushBack(frequency)
|
|
nc := fn.COUPNUM(fnArgs)
|
|
lastCoupon := firstCoupon.Number
|
|
aggrFunc := func(acc []float64, index float64) []float64 {
|
|
lastCouponTime := timeFromExcelTime(lastCoupon, false)
|
|
earlyCoupon := daysBetween(excelMinTime1900.Unix(), makeDate(lastCouponTime.Year(), time.Month(float64(lastCouponTime.Month())+numMonthsNeg), lastCouponTime.Day())) + 1
|
|
earlyCouponTime := timeFromExcelTime(earlyCoupon, false)
|
|
nl := e.Number
|
|
if basis == 1 {
|
|
nl = coupdays(earlyCouponTime, lastCouponTime, basis)
|
|
}
|
|
dci := coupdays(issueTime, lastCouponTime, basis)
|
|
if index > 1 {
|
|
dci = nl
|
|
}
|
|
startDate := earlyCoupon
|
|
if issue.Number > earlyCoupon {
|
|
startDate = issue.Number
|
|
}
|
|
endDate := lastCoupon
|
|
if settlement.Number < lastCoupon {
|
|
endDate = settlement.Number
|
|
}
|
|
startDateTime := timeFromExcelTime(startDate, false)
|
|
endDateTime := timeFromExcelTime(endDate, false)
|
|
a := coupdays(startDateTime, endDateTime, basis)
|
|
lastCoupon = earlyCoupon
|
|
dcnl := acc[0]
|
|
anl := acc[1]
|
|
return []float64{dcnl + dci/nl, anl + a/nl}
|
|
}
|
|
ag := aggrBetween(math.Floor(nc.Number), 1, []float64{0, 0}, aggrFunc)
|
|
dcnl, anl := ag[0], ag[1]
|
|
dsc := 0.0
|
|
fnArgs.Init()
|
|
fnArgs.PushBack(settlement)
|
|
fnArgs.PushBack(firstCoupon)
|
|
fnArgs.PushBack(frequency)
|
|
if basis == 2 || basis == 3 {
|
|
d := timeFromExcelTime(fn.COUPNCD(fnArgs).Number, false)
|
|
dsc = coupdays(settlementTime, d, basis)
|
|
} else {
|
|
d := timeFromExcelTime(fn.COUPPCD(fnArgs).Number, false)
|
|
a := coupdays(d, settlementTime, basis)
|
|
dsc = e.Number - a
|
|
}
|
|
nq := coupNumber(firstCoupon.Number, settlement.Number, numMonths)
|
|
fnArgs.Init()
|
|
fnArgs.PushBack(firstCoupon)
|
|
fnArgs.PushBack(maturity)
|
|
fnArgs.PushBack(frequency)
|
|
fnArgs.PushBack(basisArg)
|
|
n = fn.COUPNUM(fnArgs)
|
|
x := yld.Number/m + 1
|
|
y := dsc / e.Number
|
|
p1 := x
|
|
p3 := math.Pow(p1, y+nq+n.Number)
|
|
term1 := redemption.Number / p3
|
|
term2 := 100 * rate.Number / m * dcnl / math.Pow(p1, nq+y)
|
|
f := func(acc []float64, index float64) []float64 {
|
|
return []float64{acc[0] + 100*rate.Number/m/math.Pow(p1, index+nq+y)}
|
|
}
|
|
term3 := aggrBetween(1, math.Floor(n.Number), []float64{0}, f)
|
|
term4 := 100 * rate.Number / m * anl
|
|
return newNumberFormulaArg(term1 + term2 + term3[0] - term4)
|
|
}
|
|
|
|
// PDURATION function calculates the number of periods required for an
|
|
// investment to reach a specified future value. The syntax of the function
|
|
// is:
|
|
//
|
|
// PDURATION(rate,pv,fv)
|
|
func (fn *formulaFuncs) PDURATION(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PDURATION requires 3 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
pv := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
fv := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
if rate.Number <= 0 || pv.Number <= 0 || fv.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg((math.Log(fv.Number) - math.Log(pv.Number)) / math.Log(1+rate.Number))
|
|
}
|
|
|
|
// PMT function calculates the constant periodic payment required to pay off
|
|
// (or partially pay off) a loan or investment, with a constant interest
|
|
// rate, over a specified period. The syntax of the function is:
|
|
//
|
|
// PMT(rate,nper,pv,[fv],[type])
|
|
func (fn *formulaFuncs) PMT(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PMT requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PMT allows at most 5 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
nper := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pv := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
fv, typ := newNumberFormulaArg(0), newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if fv = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
}
|
|
if argsList.Len() == 5 {
|
|
if typ = argsList.Back().Value.(formulaArg).ToNumber(); typ.Type != ArgNumber {
|
|
return typ
|
|
}
|
|
}
|
|
if typ.Number != 0 && typ.Number != 1 {
|
|
return newErrorFormulaArg(formulaErrorNA, formulaErrorNA)
|
|
}
|
|
if rate.Number != 0 {
|
|
p := (-fv.Number - pv.Number*math.Pow(1+rate.Number, nper.Number)) / (1 + rate.Number*typ.Number) / ((math.Pow(1+rate.Number, nper.Number) - 1) / rate.Number)
|
|
return newNumberFormulaArg(p)
|
|
}
|
|
return newNumberFormulaArg((-pv.Number - fv.Number) / nper.Number)
|
|
}
|
|
|
|
// PPMT function calculates the payment on the principal, during a specific
|
|
// period of a loan or investment that is paid in constant periodic payments,
|
|
// with a constant interest rate. The syntax of the function is:
|
|
//
|
|
// PPMT(rate,per,nper,pv,[fv],[type])
|
|
func (fn *formulaFuncs) PPMT(argsList *list.List) formulaArg {
|
|
return fn.ipmt("PPMT", argsList)
|
|
}
|
|
|
|
// price is an implementation of the formula function PRICE.
|
|
func (fn *formulaFuncs) price(settlement, maturity, rate, yld, redemption, frequency, basis formulaArg) formulaArg {
|
|
if basis.Number < 0 || basis.Number > 4 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "invalid basis")
|
|
}
|
|
argsList := list.New().Init()
|
|
argsList.PushBack(settlement)
|
|
argsList.PushBack(maturity)
|
|
argsList.PushBack(frequency)
|
|
argsList.PushBack(basis)
|
|
e := fn.COUPDAYS(argsList)
|
|
dsc := fn.COUPDAYSNC(argsList).Number / e.Number
|
|
n := fn.COUPNUM(argsList)
|
|
a := fn.COUPDAYBS(argsList)
|
|
ret := 0.0
|
|
if n.Number > 1 {
|
|
ret = redemption.Number / math.Pow(1+yld.Number/frequency.Number, n.Number-1+dsc)
|
|
ret -= 100 * rate.Number / frequency.Number * a.Number / e.Number
|
|
t1 := 100 * rate.Number / frequency.Number
|
|
t2 := 1 + yld.Number/frequency.Number
|
|
for k := 0.0; k < n.Number; k++ {
|
|
ret += t1 / math.Pow(t2, k+dsc)
|
|
}
|
|
} else {
|
|
dsc = e.Number - a.Number
|
|
t1 := 100*(rate.Number/frequency.Number) + redemption.Number
|
|
t2 := (yld.Number/frequency.Number)*(dsc/e.Number) + 1
|
|
t3 := 100 * (rate.Number / frequency.Number) * (a.Number / e.Number)
|
|
ret = t1/t2 - t3
|
|
}
|
|
return newNumberFormulaArg(ret)
|
|
}
|
|
|
|
// PRICE function calculates the price, per $100 face value of a security that
|
|
// pays periodic interest. The syntax of the function is:
|
|
//
|
|
// PRICE(settlement,maturity,rate,yld,redemption,frequency,[basis])
|
|
func (fn *formulaFuncs) PRICE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 6 && argsList.Len() != 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PRICE requires 6 or 7 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
rate := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires rate >= 0")
|
|
}
|
|
yld := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if yld.Type != ArgNumber {
|
|
return yld
|
|
}
|
|
if yld.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires yld >= 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return redemption
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires redemption > 0")
|
|
}
|
|
frequency := argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if frequency.Type != ArgNumber {
|
|
return frequency
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 7 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return fn.price(settlement, maturity, rate, yld, redemption, frequency, basis)
|
|
}
|
|
|
|
// PRICEDISC function calculates the price, per $100 face value of a
|
|
// discounted security. The syntax of the function is:
|
|
//
|
|
// PRICEDISC(settlement,maturity,discount,redemption,[basis])
|
|
func (fn *formulaFuncs) PRICEDISC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PRICEDISC requires 4 or 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
if maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEDISC requires maturity > settlement")
|
|
}
|
|
discount := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if discount.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if discount.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEDISC requires discount > 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEDISC requires redemption > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg(redemption.Number * (1 - discount.Number*frac.Number))
|
|
}
|
|
|
|
// PRICEMAT function calculates the price, per $100 face value of a security
|
|
// that pays interest at maturity. The syntax of the function is:
|
|
//
|
|
// PRICEMAT(settlement,maturity,issue,rate,yld,[basis])
|
|
func (fn *formulaFuncs) PRICEMAT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 5 && argsList.Len() != 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PRICEMAT requires 5 or 6 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(3, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity, issue := args.List[0], args.List[1], args.List[2]
|
|
if settlement.Number >= maturity.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEMAT requires maturity > settlement")
|
|
}
|
|
if issue.Number >= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEMAT requires settlement > issue")
|
|
}
|
|
rate := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEMAT requires rate >= 0")
|
|
}
|
|
yld := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if yld.Type != ArgNumber {
|
|
return yld
|
|
}
|
|
if yld.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICEMAT requires yld >= 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 6 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
dsm := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if dsm.Type != ArgNumber {
|
|
return dsm
|
|
}
|
|
dis := yearFrac(issue.Number, settlement.Number, int(basis.Number))
|
|
dim := yearFrac(issue.Number, maturity.Number, int(basis.Number))
|
|
return newNumberFormulaArg(((1+dim.Number*rate.Number)/(1+dsm.Number*yld.Number) - dis.Number*rate.Number) * 100)
|
|
}
|
|
|
|
// PV function calculates the Present Value of an investment, based on a
|
|
// series of future payments. The syntax of the function is:
|
|
//
|
|
// PV(rate,nper,pmt,[fv],[type])
|
|
func (fn *formulaFuncs) PV(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PV requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "PV allows at most 5 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
nper := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pmt := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pmt.Type != ArgNumber {
|
|
return pmt
|
|
}
|
|
fv := newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if fv = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
}
|
|
t := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if t = argsList.Back().Value.(formulaArg).ToNumber(); t.Type != ArgNumber {
|
|
return t
|
|
}
|
|
if t.Number != 0 {
|
|
t.Number = 1
|
|
}
|
|
}
|
|
if rate.Number == 0 {
|
|
return newNumberFormulaArg(-pmt.Number*nper.Number - fv.Number)
|
|
}
|
|
return newNumberFormulaArg((((1-math.Pow(1+rate.Number, nper.Number))/rate.Number)*pmt.Number*(1+rate.Number*t.Number) - fv.Number) / math.Pow(1+rate.Number, nper.Number))
|
|
}
|
|
|
|
// rate is an implementation of the formula function RATE.
|
|
func (fn *formulaFuncs) rate(nper, pmt, pv, fv, t, guess formulaArg) formulaArg {
|
|
maxIter, iter, isClose, epsMax, rate := 100, 0, false, 1e-6, guess.Number
|
|
for iter < maxIter && !isClose {
|
|
t1 := math.Pow(rate+1, nper.Number)
|
|
t2 := math.Pow(rate+1, nper.Number-1)
|
|
rt := rate*t.Number + 1
|
|
p0 := pmt.Number * (t1 - 1)
|
|
f1 := fv.Number + t1*pv.Number + p0*rt/rate
|
|
n1 := nper.Number * t2 * pv.Number
|
|
n2 := p0 * rt / math.Pow(rate, 2)
|
|
f2 := math.Nextafter(n1, n1) - math.Nextafter(n2, n2)
|
|
f3 := (nper.Number*pmt.Number*t2*rt + p0*t.Number) / rate
|
|
delta := f1 / (f2 + f3)
|
|
if math.Abs(delta) < epsMax {
|
|
isClose = true
|
|
}
|
|
iter++
|
|
rate -= delta
|
|
}
|
|
return newNumberFormulaArg(rate)
|
|
}
|
|
|
|
// RATE function calculates the interest rate required to pay off a specified
|
|
// amount of a loan, or to reach a target amount on an investment, over a
|
|
// given period. The syntax of the function is:
|
|
//
|
|
// RATE(nper,pmt,pv,[fv],[type],[guess])
|
|
func (fn *formulaFuncs) RATE(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RATE requires at least 3 arguments")
|
|
}
|
|
if argsList.Len() > 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RATE allows at most 6 arguments")
|
|
}
|
|
nper := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber {
|
|
return nper
|
|
}
|
|
pmt := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if pmt.Type != ArgNumber {
|
|
return pmt
|
|
}
|
|
pv := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pv.Type != ArgNumber {
|
|
return pv
|
|
}
|
|
fv := newNumberFormulaArg(0)
|
|
if argsList.Len() >= 4 {
|
|
if fv = argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber(); fv.Type != ArgNumber {
|
|
return fv
|
|
}
|
|
}
|
|
t := newNumberFormulaArg(0)
|
|
if argsList.Len() >= 5 {
|
|
if t = argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); t.Type != ArgNumber {
|
|
return t
|
|
}
|
|
if t.Number != 0 {
|
|
t.Number = 1
|
|
}
|
|
}
|
|
guess := newNumberFormulaArg(0.1)
|
|
if argsList.Len() == 6 {
|
|
if guess = argsList.Back().Value.(formulaArg).ToNumber(); guess.Type != ArgNumber {
|
|
return guess
|
|
}
|
|
}
|
|
return fn.rate(nper, pmt, pv, fv, t, guess)
|
|
}
|
|
|
|
// RECEIVED function calculates the amount received at maturity for a fully
|
|
// invested security. The syntax of the function is:
|
|
//
|
|
// RECEIVED(settlement,maturity,investment,discount,[basis])
|
|
func (fn *formulaFuncs) RECEIVED(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RECEIVED requires at least 4 arguments")
|
|
}
|
|
if argsList.Len() > 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RECEIVED allows at most 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
investment := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if investment.Type != ArgNumber {
|
|
return investment
|
|
}
|
|
discount := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if discount.Type != ArgNumber {
|
|
return discount
|
|
}
|
|
if discount.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "RECEIVED requires discount > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg(investment.Number / (1 - discount.Number*frac.Number))
|
|
}
|
|
|
|
// RRI function calculates the equivalent interest rate for an investment with
|
|
// specified present value, future value and duration. The syntax of the
|
|
// function is:
|
|
//
|
|
// RRI(nper,pv,fv)
|
|
func (fn *formulaFuncs) RRI(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "RRI requires 3 arguments")
|
|
}
|
|
nper := argsList.Front().Value.(formulaArg).ToNumber()
|
|
pv := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
fv := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if nper.Type != ArgNumber || pv.Type != ArgNumber || fv.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if nper.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "RRI requires nper argument to be > 0")
|
|
}
|
|
if pv.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "RRI requires pv argument to be > 0")
|
|
}
|
|
if fv.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "RRI requires fv argument to be >= 0")
|
|
}
|
|
return newNumberFormulaArg(math.Pow(fv.Number/pv.Number, 1/nper.Number) - 1)
|
|
}
|
|
|
|
// SLN function calculates the straight line depreciation of an asset for one
|
|
// period. The syntax of the function is:
|
|
//
|
|
// SLN(cost,salvage,life)
|
|
func (fn *formulaFuncs) SLN(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SLN requires 3 arguments")
|
|
}
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
salvage := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
life := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber || salvage.Type != ArgNumber || life.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if life.Number == 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "SLN requires life argument to be > 0")
|
|
}
|
|
return newNumberFormulaArg((cost.Number - salvage.Number) / life.Number)
|
|
}
|
|
|
|
// SYD function calculates the sum-of-years' digits depreciation for a
|
|
// specified period in the lifetime of an asset. The syntax of the function
|
|
// is:
|
|
//
|
|
// SYD(cost,salvage,life,per)
|
|
func (fn *formulaFuncs) SYD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "SYD requires 4 arguments")
|
|
}
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
salvage := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
life := argsList.Back().Prev().Value.(formulaArg).ToNumber()
|
|
per := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber || salvage.Type != ArgNumber || life.Type != ArgNumber || per.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if life.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "SYD requires life argument to be > 0")
|
|
}
|
|
if per.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "SYD requires per argument to be > 0")
|
|
}
|
|
if per.Number > life.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(((cost.Number - salvage.Number) * (life.Number - per.Number + 1) * 2) / (life.Number * (life.Number + 1)))
|
|
}
|
|
|
|
// TBILLEQ function calculates the bond-equivalent yield for a Treasury Bill.
|
|
// The syntax of the function is:
|
|
//
|
|
// TBILLEQ(settlement,maturity,discount)
|
|
func (fn *formulaFuncs) TBILLEQ(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TBILLEQ requires 3 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
dsm := maturity.Number - settlement.Number
|
|
if dsm > 365 || maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
discount := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if discount.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if discount.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg((365 * discount.Number) / (360 - discount.Number*dsm))
|
|
}
|
|
|
|
// TBILLPRICE function returns the price, per $100 face value, of a Treasury
|
|
// Bill. The syntax of the function is:
|
|
//
|
|
// TBILLPRICE(settlement,maturity,discount)
|
|
func (fn *formulaFuncs) TBILLPRICE(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TBILLPRICE requires 3 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
dsm := maturity.Number - settlement.Number
|
|
if dsm > 365 || maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
discount := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if discount.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if discount.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(100 * (1 - discount.Number*dsm/360))
|
|
}
|
|
|
|
// TBILLYIELD function calculates the yield of a Treasury Bill. The syntax of
|
|
// the function is:
|
|
//
|
|
// TBILLYIELD(settlement,maturity,pr)
|
|
func (fn *formulaFuncs) TBILLYIELD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "TBILLYIELD requires 3 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
dsm := maturity.Number - settlement.Number
|
|
if dsm > 365 || maturity.Number <= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
pr := argsList.Back().Value.(formulaArg).ToNumber()
|
|
if pr.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if pr.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(((100 - pr.Number) / pr.Number) * (360 / dsm))
|
|
}
|
|
|
|
// prepareVdbArgs checking and prepare arguments for the formula function
|
|
// VDB.
|
|
func (fn *formulaFuncs) prepareVdbArgs(argsList *list.List) formulaArg {
|
|
cost := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if cost.Type != ArgNumber {
|
|
return cost
|
|
}
|
|
if cost.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires cost >= 0")
|
|
}
|
|
salvage := argsList.Front().Next().Value.(formulaArg).ToNumber()
|
|
if salvage.Type != ArgNumber {
|
|
return salvage
|
|
}
|
|
if salvage.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires salvage >= 0")
|
|
}
|
|
life := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if life.Type != ArgNumber {
|
|
return life
|
|
}
|
|
if life.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires life > 0")
|
|
}
|
|
startPeriod := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if startPeriod.Type != ArgNumber {
|
|
return startPeriod
|
|
}
|
|
if startPeriod.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires start_period > 0")
|
|
}
|
|
endPeriod := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if endPeriod.Type != ArgNumber {
|
|
return endPeriod
|
|
}
|
|
if startPeriod.Number > endPeriod.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires start_period <= end_period")
|
|
}
|
|
if endPeriod.Number > life.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "VDB requires end_period <= life")
|
|
}
|
|
factor := newNumberFormulaArg(2)
|
|
if argsList.Len() > 5 {
|
|
if factor = argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber(); factor.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if factor.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "VDB requires factor >= 0")
|
|
}
|
|
}
|
|
return newListFormulaArg([]formulaArg{cost, salvage, life, startPeriod, endPeriod, factor})
|
|
}
|
|
|
|
// vdb is a part of implementation of the formula function VDB.
|
|
func (fn *formulaFuncs) vdb(cost, salvage, life, life1, period, factor formulaArg) formulaArg {
|
|
var ddb, vdb, sln, term float64
|
|
endInt, cs, nowSln := math.Ceil(period.Number), cost.Number-salvage.Number, false
|
|
ddbArgs := list.New()
|
|
for i := 1.0; i <= endInt; i++ {
|
|
if !nowSln {
|
|
ddbArgs.Init()
|
|
ddbArgs.PushBack(cost)
|
|
ddbArgs.PushBack(salvage)
|
|
ddbArgs.PushBack(life)
|
|
ddbArgs.PushBack(newNumberFormulaArg(i))
|
|
ddbArgs.PushBack(factor)
|
|
ddb = fn.DDB(ddbArgs).Number
|
|
sln = cs / (life1.Number - i + 1)
|
|
if sln > ddb {
|
|
term = sln
|
|
nowSln = true
|
|
} else {
|
|
term = ddb
|
|
cs -= ddb
|
|
}
|
|
} else {
|
|
term = sln
|
|
}
|
|
if i == endInt {
|
|
term *= period.Number + 1 - endInt
|
|
}
|
|
vdb += term
|
|
}
|
|
return newNumberFormulaArg(vdb)
|
|
}
|
|
|
|
// VDB function calculates the depreciation of an asset, using the Double
|
|
// Declining Balance Method, or another specified depreciation rate, for a
|
|
// specified period (including partial periods). The syntax of the function
|
|
// is:
|
|
//
|
|
// VDB(cost,salvage,life,start_period,end_period,[factor],[no_switch])
|
|
func (fn *formulaFuncs) VDB(argsList *list.List) formulaArg {
|
|
if argsList.Len() < 5 || argsList.Len() > 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "VDB requires 5 or 7 arguments")
|
|
}
|
|
args := fn.prepareVdbArgs(argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
cost, salvage, life, startPeriod, endPeriod, factor := args.List[0], args.List[1], args.List[2], args.List[3], args.List[4], args.List[5]
|
|
noSwitch := newBoolFormulaArg(false)
|
|
if argsList.Len() > 6 {
|
|
if noSwitch = argsList.Back().Value.(formulaArg).ToBool(); noSwitch.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
startInt, endInt, vdb, ddbArgs := math.Floor(startPeriod.Number), math.Ceil(endPeriod.Number), newNumberFormulaArg(0), list.New()
|
|
if noSwitch.Number == 1 {
|
|
for i := startInt + 1; i <= endInt; i++ {
|
|
ddbArgs.Init()
|
|
ddbArgs.PushBack(cost)
|
|
ddbArgs.PushBack(salvage)
|
|
ddbArgs.PushBack(life)
|
|
ddbArgs.PushBack(newNumberFormulaArg(i))
|
|
ddbArgs.PushBack(factor)
|
|
term := fn.DDB(ddbArgs)
|
|
if i == startInt+1 {
|
|
term.Number *= math.Min(endPeriod.Number, startInt+1) - startPeriod.Number
|
|
} else if i == endInt {
|
|
term.Number *= endPeriod.Number + 1 - endInt
|
|
}
|
|
vdb.Number += term.Number
|
|
}
|
|
return vdb
|
|
}
|
|
life1, part := life, 0.0
|
|
if startPeriod.Number != math.Floor(startPeriod.Number) && factor.Number > 1.0 && startPeriod.Number >= life.Number/2.0 {
|
|
part = startPeriod.Number - life.Number/2.0
|
|
startPeriod.Number = life.Number / 2.0
|
|
endPeriod.Number -= part
|
|
}
|
|
cost.Number -= fn.vdb(cost, salvage, life, life1, startPeriod, factor).Number
|
|
return fn.vdb(cost, salvage, life, newNumberFormulaArg(life.Number-startPeriod.Number), newNumberFormulaArg(endPeriod.Number-startPeriod.Number), factor)
|
|
}
|
|
|
|
// prepareXArgs prepare arguments for the formula function XIRR and XNPV.
|
|
func (fn *formulaFuncs) prepareXArgs(values, dates formulaArg) (valuesArg, datesArg []float64, err formulaArg) {
|
|
for _, arg := range values.ToList() {
|
|
if numArg := arg.ToNumber(); numArg.Type == ArgNumber {
|
|
valuesArg = append(valuesArg, numArg.Number)
|
|
continue
|
|
}
|
|
err = newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
return
|
|
}
|
|
if len(valuesArg) < 2 {
|
|
err = newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
return
|
|
}
|
|
args, date := list.New(), 0.0
|
|
for _, arg := range dates.ToList() {
|
|
args.Init()
|
|
args.PushBack(arg)
|
|
dateValue := fn.DATEVALUE(args)
|
|
if dateValue.Type != ArgNumber {
|
|
err = newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
return
|
|
}
|
|
if dateValue.Number < date {
|
|
err = newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
return
|
|
}
|
|
datesArg = append(datesArg, dateValue.Number)
|
|
date = dateValue.Number
|
|
}
|
|
if len(valuesArg) != len(datesArg) {
|
|
err = newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
return
|
|
}
|
|
err = newEmptyFormulaArg()
|
|
return
|
|
}
|
|
|
|
// xirr is an implementation of the formula function XIRR.
|
|
func (fn *formulaFuncs) xirr(values, dates []float64, guess float64) formulaArg {
|
|
positive, negative := false, false
|
|
for i := 0; i < len(values); i++ {
|
|
if values[i] > 0 {
|
|
positive = true
|
|
}
|
|
if values[i] < 0 {
|
|
negative = true
|
|
}
|
|
}
|
|
if !positive || !negative {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
result, epsMax, count, maxIterate, err := guess, 1e-10, 0, 50, false
|
|
for {
|
|
resultValue := xirrPart1(values, dates, result)
|
|
newRate := result - resultValue/xirrPart2(values, dates, result)
|
|
epsRate := math.Abs(newRate - result)
|
|
result = newRate
|
|
count++
|
|
if epsRate <= epsMax || math.Abs(resultValue) <= epsMax {
|
|
break
|
|
}
|
|
if count > maxIterate {
|
|
err = true
|
|
break
|
|
}
|
|
}
|
|
if err || math.IsNaN(result) || math.IsInf(result, 0) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// xirrPart1 is a part of implementation of the formula function XIRR.
|
|
func xirrPart1(values, dates []float64, rate float64) float64 {
|
|
r := rate + 1
|
|
result := values[0]
|
|
vlen := len(values)
|
|
firstDate := dates[0]
|
|
for i := 1; i < vlen; i++ {
|
|
result += values[i] / math.Pow(r, (dates[i]-firstDate)/365)
|
|
}
|
|
return result
|
|
}
|
|
|
|
// xirrPart2 is a part of implementation of the formula function XIRR.
|
|
func xirrPart2(values, dates []float64, rate float64) float64 {
|
|
r := rate + 1
|
|
result := 0.0
|
|
vlen := len(values)
|
|
firstDate := dates[0]
|
|
for i := 1; i < vlen; i++ {
|
|
frac := (dates[i] - firstDate) / 365
|
|
result -= frac * values[i] / math.Pow(r, frac+1)
|
|
}
|
|
return result
|
|
}
|
|
|
|
// XIRR function returns the Internal Rate of Return for a supplied series of
|
|
// cash flows (i.e. a set of values, which includes an initial investment
|
|
// value and a series of net income values) occurring at a series of supplied
|
|
// dates. The syntax of the function is:
|
|
//
|
|
// XIRR(values,dates,[guess])
|
|
func (fn *formulaFuncs) XIRR(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 2 && argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XIRR requires 2 or 3 arguments")
|
|
}
|
|
values, dates, err := fn.prepareXArgs(argsList.Front().Value.(formulaArg), argsList.Front().Next().Value.(formulaArg))
|
|
if err.Type != ArgEmpty {
|
|
return err
|
|
}
|
|
guess := newNumberFormulaArg(0)
|
|
if argsList.Len() == 3 {
|
|
if guess = argsList.Back().Value.(formulaArg).ToNumber(); guess.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
if guess.Number <= -1 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XIRR requires guess > -1")
|
|
}
|
|
}
|
|
return fn.xirr(values, dates, guess.Number)
|
|
}
|
|
|
|
// XNPV function calculates the Net Present Value for a schedule of cash flows
|
|
// that is not necessarily periodic. The syntax of the function is:
|
|
//
|
|
// XNPV(rate,values,dates)
|
|
func (fn *formulaFuncs) XNPV(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XNPV requires 3 arguments")
|
|
}
|
|
rate := argsList.Front().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "XNPV requires rate > 0")
|
|
}
|
|
values, dates, err := fn.prepareXArgs(argsList.Front().Next().Value.(formulaArg), argsList.Back().Value.(formulaArg))
|
|
if err.Type != ArgEmpty {
|
|
return err
|
|
}
|
|
date1, xnpv := dates[0], 0.0
|
|
for idx, value := range values {
|
|
xnpv += value / math.Pow(1+rate.Number, (dates[idx]-date1)/365)
|
|
}
|
|
return newNumberFormulaArg(xnpv)
|
|
}
|
|
|
|
// yield is an implementation of the formula function YIELD.
|
|
func (fn *formulaFuncs) yield(settlement, maturity, rate, pr, redemption, frequency, basis formulaArg) formulaArg {
|
|
priceN, yield1, yield2 := newNumberFormulaArg(0), newNumberFormulaArg(0), newNumberFormulaArg(1)
|
|
price1 := fn.price(settlement, maturity, rate, yield1, redemption, frequency, basis)
|
|
if price1.Type != ArgNumber {
|
|
return price1
|
|
}
|
|
price2 := fn.price(settlement, maturity, rate, yield2, redemption, frequency, basis)
|
|
yieldN := newNumberFormulaArg((yield2.Number - yield1.Number) * 0.5)
|
|
for iter := 0; iter < 100 && priceN.Number != pr.Number; iter++ {
|
|
priceN = fn.price(settlement, maturity, rate, yieldN, redemption, frequency, basis)
|
|
if pr.Number == price1.Number {
|
|
return yield1
|
|
} else if pr.Number == price2.Number {
|
|
return yield2
|
|
} else if pr.Number == priceN.Number {
|
|
return yieldN
|
|
} else if pr.Number < price2.Number {
|
|
yield2.Number *= 2.0
|
|
price2 = fn.price(settlement, maturity, rate, yield2, redemption, frequency, basis)
|
|
yieldN.Number = (yield2.Number - yield1.Number) * 0.5
|
|
} else {
|
|
if pr.Number < priceN.Number {
|
|
yield1 = yieldN
|
|
price1 = priceN
|
|
} else {
|
|
yield2 = yieldN
|
|
price2 = priceN
|
|
}
|
|
f1 := (yield2.Number - yield1.Number) * ((pr.Number - price2.Number) / (price1.Number - price2.Number))
|
|
yieldN.Number = yield2.Number - math.Nextafter(f1, f1)
|
|
}
|
|
}
|
|
return yieldN
|
|
}
|
|
|
|
// YIELD function calculates the Yield of a security that pays periodic
|
|
// interest. The syntax of the function is:
|
|
//
|
|
// YIELD(settlement,maturity,rate,pr,redemption,frequency,[basis])
|
|
func (fn *formulaFuncs) YIELD(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 6 && argsList.Len() != 7 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "YIELD requires 6 or 7 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
rate := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires rate >= 0")
|
|
}
|
|
pr := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pr.Type != ArgNumber {
|
|
return pr
|
|
}
|
|
if pr.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires pr > 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return redemption
|
|
}
|
|
if redemption.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "PRICE requires redemption >= 0")
|
|
}
|
|
frequency := argsList.Front().Next().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if frequency.Type != ArgNumber {
|
|
return frequency
|
|
}
|
|
if !validateFrequency(frequency.Number) {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 7 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return fn.yield(settlement, maturity, rate, pr, redemption, frequency, basis)
|
|
}
|
|
|
|
// YIELDDISC function calculates the annual yield of a discounted security.
|
|
// The syntax of the function is:
|
|
//
|
|
// YIELDDISC(settlement,maturity,pr,redemption,[basis])
|
|
func (fn *formulaFuncs) YIELDDISC(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 4 && argsList.Len() != 5 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "YIELDDISC requires 4 or 5 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
pr := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pr.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if pr.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YIELDDISC requires pr > 0")
|
|
}
|
|
redemption := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if redemption.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
if redemption.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YIELDDISC requires redemption > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 5 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
frac := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
if frac.Type != ArgNumber {
|
|
return frac
|
|
}
|
|
return newNumberFormulaArg((redemption.Number/pr.Number - 1) / frac.Number)
|
|
}
|
|
|
|
// YIELDMAT function calculates the annual yield of a security that pays
|
|
// interest at maturity. The syntax of the function is:
|
|
//
|
|
// YIELDMAT(settlement,maturity,issue,rate,pr,[basis])
|
|
func (fn *formulaFuncs) YIELDMAT(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 5 && argsList.Len() != 6 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "YIELDMAT requires 5 or 6 arguments")
|
|
}
|
|
args := fn.prepareDataValueArgs(2, argsList)
|
|
if args.Type != ArgList {
|
|
return args
|
|
}
|
|
settlement, maturity := args.List[0], args.List[1]
|
|
arg := list.New().Init()
|
|
issue := argsList.Front().Next().Next().Value.(formulaArg).ToNumber()
|
|
if issue.Type != ArgNumber {
|
|
arg.PushBack(argsList.Front().Next().Next().Value.(formulaArg))
|
|
issue = fn.DATEVALUE(arg)
|
|
if issue.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
}
|
|
if issue.Number >= settlement.Number {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YIELDMAT requires settlement > issue")
|
|
}
|
|
rate := argsList.Front().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if rate.Type != ArgNumber {
|
|
return rate
|
|
}
|
|
if rate.Number < 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YIELDMAT requires rate >= 0")
|
|
}
|
|
pr := argsList.Front().Next().Next().Next().Next().Value.(formulaArg).ToNumber()
|
|
if pr.Type != ArgNumber {
|
|
return pr
|
|
}
|
|
if pr.Number <= 0 {
|
|
return newErrorFormulaArg(formulaErrorNUM, "YIELDMAT requires pr > 0")
|
|
}
|
|
basis := newNumberFormulaArg(0)
|
|
if argsList.Len() == 6 {
|
|
if basis = argsList.Back().Value.(formulaArg).ToNumber(); basis.Type != ArgNumber {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
dim := yearFrac(issue.Number, maturity.Number, int(basis.Number))
|
|
if dim.Type != ArgNumber {
|
|
return dim
|
|
}
|
|
dis := yearFrac(issue.Number, settlement.Number, int(basis.Number))
|
|
dsm := yearFrac(settlement.Number, maturity.Number, int(basis.Number))
|
|
f1 := dim.Number * rate.Number
|
|
result := 1 + math.Nextafter(f1, f1)
|
|
result /= pr.Number/100 + dis.Number*rate.Number
|
|
result--
|
|
result /= dsm.Number
|
|
return newNumberFormulaArg(result)
|
|
}
|
|
|
|
// Database Functions
|
|
|
|
// calcDatabase defines the structure for formula database.
|
|
type calcDatabase struct {
|
|
col, row int
|
|
indexMap map[int]int
|
|
database [][]formulaArg
|
|
criteria [][]formulaArg
|
|
}
|
|
|
|
// newCalcDatabase function returns formula database by given data range of
|
|
// cells containing the database, field and criteria range.
|
|
func newCalcDatabase(database, field, criteria formulaArg) *calcDatabase {
|
|
db := calcDatabase{
|
|
indexMap: make(map[int]int),
|
|
database: database.Matrix,
|
|
criteria: criteria.Matrix,
|
|
}
|
|
exp := len(database.Matrix) < 2 || len(database.Matrix[0]) < 1 ||
|
|
len(criteria.Matrix) < 2 || len(criteria.Matrix[0]) < 1
|
|
if field.Type != ArgEmpty {
|
|
if db.col = db.columnIndex(database.Matrix, field); exp || db.col < 0 || len(db.database[0]) <= db.col {
|
|
return nil
|
|
}
|
|
return &db
|
|
}
|
|
if db.col = -1; exp {
|
|
return nil
|
|
}
|
|
return &db
|
|
}
|
|
|
|
// columnIndex return index by specifies column field within the database for
|
|
// which user want to return the count of non-blank cells.
|
|
func (db *calcDatabase) columnIndex(database [][]formulaArg, field formulaArg) int {
|
|
num := field.ToNumber()
|
|
if num.Type != ArgNumber && len(database) > 0 {
|
|
for i := 0; i < len(database[0]); i++ {
|
|
if title := database[0][i]; strings.EqualFold(title.Value(), field.Value()) {
|
|
return i
|
|
}
|
|
}
|
|
return -1
|
|
}
|
|
return int(num.Number - 1)
|
|
}
|
|
|
|
// criteriaEval evaluate formula criteria expression.
|
|
func (db *calcDatabase) criteriaEval() bool {
|
|
var (
|
|
columns, rows = len(db.criteria[0]), len(db.criteria)
|
|
criteria = db.criteria
|
|
k int
|
|
matched bool
|
|
)
|
|
if len(db.indexMap) == 0 {
|
|
fields := criteria[0]
|
|
for j := 0; j < columns; j++ {
|
|
if k = db.columnIndex(db.database, fields[j]); k < 0 {
|
|
return false
|
|
}
|
|
db.indexMap[j] = k
|
|
}
|
|
}
|
|
for i := 1; !matched && i < rows; i++ {
|
|
matched = true
|
|
for j := 0; matched && j < columns; j++ {
|
|
criteriaExp := db.criteria[i][j].Value()
|
|
if criteriaExp == "" {
|
|
continue
|
|
}
|
|
criteria := formulaCriteriaParser(criteriaExp)
|
|
cell := db.database[db.row][db.indexMap[j]].Value()
|
|
matched, _ = formulaCriteriaEval(cell, criteria)
|
|
}
|
|
}
|
|
return matched
|
|
}
|
|
|
|
// value returns the current cell value.
|
|
func (db *calcDatabase) value() formulaArg {
|
|
if db.col == -1 {
|
|
return db.database[db.row][len(db.database[db.row])-1]
|
|
}
|
|
return db.database[db.row][db.col]
|
|
}
|
|
|
|
// next will return true if find the matched cell in the database.
|
|
func (db *calcDatabase) next() bool {
|
|
matched, rows := false, len(db.database)
|
|
for !matched && db.row < rows {
|
|
if db.row++; db.row < rows {
|
|
matched = db.criteriaEval()
|
|
}
|
|
}
|
|
return matched
|
|
}
|
|
|
|
// database is an implementation of the formula functions DAVERAGE, DMAX and DMIN.
|
|
func (fn *formulaFuncs) database(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires 3 arguments", name))
|
|
}
|
|
database := argsList.Front().Value.(formulaArg)
|
|
field := argsList.Front().Next().Value.(formulaArg)
|
|
criteria := argsList.Back().Value.(formulaArg)
|
|
db := newCalcDatabase(database, field, criteria)
|
|
if db == nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
args := list.New()
|
|
for db.next() {
|
|
args.PushBack(db.value())
|
|
}
|
|
switch name {
|
|
case "DMAX":
|
|
return fn.MAX(args)
|
|
case "DMIN":
|
|
return fn.MIN(args)
|
|
case "DPRODUCT":
|
|
return fn.PRODUCT(args)
|
|
case "DSTDEV":
|
|
return fn.STDEV(args)
|
|
case "DSTDEVP":
|
|
return fn.STDEVP(args)
|
|
case "DSUM":
|
|
return fn.SUM(args)
|
|
case "DVAR":
|
|
return fn.VAR(args)
|
|
case "DVARP":
|
|
return fn.VARP(args)
|
|
default:
|
|
return fn.AVERAGE(args)
|
|
}
|
|
}
|
|
|
|
// DAVERAGE function calculates the average (statistical mean) of values in a
|
|
// field (column) in a database for selected records, that satisfy
|
|
// user-specified criteria. The syntax of the Excel Daverage function is:
|
|
//
|
|
// DAVERAGE(database,field,criteria)
|
|
func (fn *formulaFuncs) DAVERAGE(argsList *list.List) formulaArg {
|
|
return fn.database("DAVERAGE", argsList)
|
|
}
|
|
|
|
// dcount is an implementation of the formula functions DCOUNT and DCOUNTA.
|
|
func (fn *formulaFuncs) dcount(name string, argsList *list.List) formulaArg {
|
|
if argsList.Len() < 2 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s requires at least 2 arguments", name))
|
|
}
|
|
if argsList.Len() > 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, fmt.Sprintf("%s allows at most 3 arguments", name))
|
|
}
|
|
field := newEmptyFormulaArg()
|
|
criteria := argsList.Back().Value.(formulaArg)
|
|
if argsList.Len() > 2 {
|
|
field = argsList.Front().Next().Value.(formulaArg)
|
|
}
|
|
database := argsList.Front().Value.(formulaArg)
|
|
db := newCalcDatabase(database, field, criteria)
|
|
if db == nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
args := list.New()
|
|
for db.next() {
|
|
args.PushBack(db.value())
|
|
}
|
|
if name == "DCOUNT" {
|
|
return fn.COUNT(args)
|
|
}
|
|
return fn.COUNTA(args)
|
|
}
|
|
|
|
// DCOUNT function returns the number of cells containing numeric values, in a
|
|
// field (column) of a database for selected records only. The records to be
|
|
// included in the count are those that satisfy a set of one or more
|
|
// user-specified criteria. The syntax of the function is:
|
|
//
|
|
// DCOUNT(database,[field],criteria)
|
|
func (fn *formulaFuncs) DCOUNT(argsList *list.List) formulaArg {
|
|
return fn.dcount("DCOUNT", argsList)
|
|
}
|
|
|
|
// DCOUNTA function returns the number of non-blank cells, in a field
|
|
// (column) of a database for selected records only. The records to be
|
|
// included in the count are those that satisfy a set of one or more
|
|
// user-specified criteria. The syntax of the function is:
|
|
//
|
|
// DCOUNTA(database,[field],criteria)
|
|
func (fn *formulaFuncs) DCOUNTA(argsList *list.List) formulaArg {
|
|
return fn.dcount("DCOUNTA", argsList)
|
|
}
|
|
|
|
// DGET function returns a single value from a column of a database. The record
|
|
// is selected via a set of one or more user-specified criteria. The syntax of
|
|
// the function is:
|
|
//
|
|
// DGET(database,field,criteria)
|
|
func (fn *formulaFuncs) DGET(argsList *list.List) formulaArg {
|
|
if argsList.Len() != 3 {
|
|
return newErrorFormulaArg(formulaErrorVALUE, "DGET requires 3 arguments")
|
|
}
|
|
database := argsList.Front().Value.(formulaArg)
|
|
field := argsList.Front().Next().Value.(formulaArg)
|
|
criteria := argsList.Back().Value.(formulaArg)
|
|
db := newCalcDatabase(database, field, criteria)
|
|
if db == nil {
|
|
return newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
}
|
|
value := newErrorFormulaArg(formulaErrorVALUE, formulaErrorVALUE)
|
|
if db.next() {
|
|
if value = db.value(); db.next() {
|
|
return newErrorFormulaArg(formulaErrorNUM, formulaErrorNUM)
|
|
}
|
|
}
|
|
return value
|
|
}
|
|
|
|
// DMAX function finds the maximum value in a field (column) in a database for
|
|
// selected records only. The records to be included in the calculation are
|
|
// defined by a set of one or more user-specified criteria. The syntax of the
|
|
// function is:
|
|
//
|
|
// DMAX(database,field,criteria)
|
|
func (fn *formulaFuncs) DMAX(argsList *list.List) formulaArg {
|
|
return fn.database("DMAX", argsList)
|
|
}
|
|
|
|
// DMIN function finds the minimum value in a field (column) in a database for
|
|
// selected records only. The records to be included in the calculation are
|
|
// defined by a set of one or more user-specified criteria. The syntax of the
|
|
// function is:
|
|
//
|
|
// DMIN(database,field,criteria)
|
|
func (fn *formulaFuncs) DMIN(argsList *list.List) formulaArg {
|
|
return fn.database("DMIN", argsList)
|
|
}
|
|
|
|
// DPRODUCT function calculates the product of a field (column) in a database
|
|
// for selected records, that satisfy user-specified criteria. The syntax of
|
|
// the function is:
|
|
//
|
|
// DPRODUCT(database,field,criteria)
|
|
func (fn *formulaFuncs) DPRODUCT(argsList *list.List) formulaArg {
|
|
return fn.database("DPRODUCT", argsList)
|
|
}
|
|
|
|
// DSTDEV function calculates the sample standard deviation of a field
|
|
// (column) in a database for selected records only. The records to be
|
|
// included in the calculation are defined by a set of one or more
|
|
// user-specified criteria. The syntax of the function is:
|
|
//
|
|
// DSTDEV(database,field,criteria)
|
|
func (fn *formulaFuncs) DSTDEV(argsList *list.List) formulaArg {
|
|
return fn.database("DSTDEV", argsList)
|
|
}
|
|
|
|
// DSTDEVP function calculates the standard deviation of a field (column) in a
|
|
// database for selected records only. The records to be included in the
|
|
// calculation are defined by a set of one or more user-specified criteria.
|
|
// The syntax of the function is:
|
|
//
|
|
// DSTDEVP(database,field,criteria)
|
|
func (fn *formulaFuncs) DSTDEVP(argsList *list.List) formulaArg {
|
|
return fn.database("DSTDEVP", argsList)
|
|
}
|
|
|
|
// DSUM function calculates the sum of a field (column) in a database for
|
|
// selected records, that satisfy user-specified criteria. The syntax of the
|
|
// function is:
|
|
//
|
|
// DSUM(database,field,criteria)
|
|
func (fn *formulaFuncs) DSUM(argsList *list.List) formulaArg {
|
|
return fn.database("DSUM", argsList)
|
|
}
|
|
|
|
// DVAR function calculates the sample variance of a field (column) in a
|
|
// database for selected records only. The records to be included in the
|
|
// calculation are defined by a set of one or more user-specified criteria.
|
|
// The syntax of the function is:
|
|
//
|
|
// DVAR(database,field,criteria)
|
|
func (fn *formulaFuncs) DVAR(argsList *list.List) formulaArg {
|
|
return fn.database("DVAR", argsList)
|
|
}
|
|
|
|
// DVARP function calculates the variance (for an entire population), of the
|
|
// values in a field (column) in a database for selected records only. The
|
|
// records to be included in the calculation are defined by a set of one or
|
|
// more user-specified criteria. The syntax of the function is:
|
|
//
|
|
// DVARP(database,field,criteria)
|
|
func (fn *formulaFuncs) DVARP(argsList *list.List) formulaArg {
|
|
return fn.database("DVARP", argsList)
|
|
}
|