convToNum: Convert Vector To Numeric

Description

This function checks if input vector/character string contains numbers (with or without comma) and attempts converting to numeric. This functions was designed for extracting the numeric part of character-vectors (or matrix) containing both numbers and character-elements. Depending on the parameters convert and remove text-entries can be converted to NA (in resulting numeric objects) or removed (the number of elements/lines gets reduced, in consequece). Note: if 'x' is a matrix, its matrix-dimensions & -names will be preserved. Note: so far Inf and -Inf do not get recognized as numeric.

Usage

convToNum(
  x,
  autoConv = TRUE,
  spaceRemove = TRUE,
  convert = c(NA, "sparseChar"),
  remove = NULL,
  euroStyle = TRUE,
  sciIncl = TRUE,
  callFrom = NULL,
  silent = TRUE,
  debug = FALSE
)

Value

This function returns a numeric vector (or matrix (if 'x' is matrix))

Arguments

x: vector to be converted
autoConv: (logical) simple automatic conversion based on function as.numeric; if TRUE all other arguments exept spaceRemove will not be considered
spaceRemove: (logical) to remove all heading and trailing (white) space (until first non-space character)
convert: (character) define which type of non-conform entries to convert to NAs. Note, if remove is selected to eliminate character-entries they cannot be converted any more. Use 'allChar' for all character-entries; 'sparseChar' sparse (ie rare) character entries; NA for converting 'Na' or 'na' to NA; if 'none' or NULL no conversions at all.
remove: (character) define which type of non-conform entries to remove, removed items cannot be converted to NA any more. Use 'allChar' for removing all character entries; NA for removing all instances of NA (execept thise created by converting text); all elements will be kept if 'none' or NULL.
euroStyle: (logical) if TRUE will convert all ',' (eg used as European decimal-separator) to '.' (as internally used by R as decimal-separator), thus allowing converting the European decimal format.
sciIncl: (logical) include recognizing scientific notation (eg 2e-4)
callFrom: (character) allow easier tracking of messages produced
silent: (logical) suppress messages
debug: (logical) additional messages for debugging

Details

This function may be used in two modes, depening if argument autoConv is TRUE or FALSE. The first options allows accessing an automatic mode based on as.numeric, while the second options investigates all characters if they may belong to numeric expressions and allows removing specific text-elements.

Examples

Run this code

x1 <- c("+4"," + 5","6","bb","Na","-7") 
convToNum(x1) 
convToNum(x1, autoConv=FALSE, convert=c("allChar"))
convToNum(x1, autoConv=FALSE)      # too many non-numeric instances for 'sparseChar'

x2 <- c("+4"," + 5","6","-7"," - 8","1e6","+ 2.3e4","-3E4","- 4E5") 
convToNum(x2) 
convToNum(x2, autoConv=FALSE, convert=NA,remove=c("allChar",NA))
convToNum(x2, autoConv=FALSE, convert=NA,remove=c("allChar",NA),sciIncl=FALSE)