R/makemeNA.R

In mrdwab/AMsnippets: Useful function "snippets"

#'Make certain values in a \code{data.frame} \code{NA}
#'
#'Sometimes, after having read in data, one needs to replace certain values by
#'\code{\link{NA}}. One approach is to use \code{mydf[mydf == "some-character"]
#'<- NA}.  However, in many cases that results in a \code{data.frame} where
#'variables which should be numeric end up as characters or factors if the NA
#'string was a character to begin with. This function is a convenience wrapper
#'around \code{\link{type.convert}} to address such problems.
#'
#'
#'@param mydf A \code{data.frame} in which some values need to be converted to
#'\code{NA}
#'@param NAStrings The values which have been used to represent \code{NA}
#'@param fixed Logical. Is the \code{NAStrings} argument a fixed character (or
#'vector of characters) or a regular expression? Defaults to \code{TRUE}.
#'@param overwrite Logical. Should the current object be overwritten? Defaults
#'to \code{FALSE}
#'@author Ananda Mahto
#'@seealso \code{\link{type.convert}}
#'@examples
#'
#'# Some sample data
#'temp <- data.frame(
#'V1 = c(1:3),
#'V2 = c(1, "*", 3),
#'V3 = c("a", "*", "c"),
#'V4 = c(".", "*", "3"))
#'temp
#'str(temp)
#'
#'temp1 <- makemeNA(temp, c("*", "."))
#'temp1
#'str(temp1)
#'
#'# Can make anything NA. Useful for -999 type of NA values
#'makemeNA(temp, "1")
#'\dontshow{rm(temp, temp1)}
#'
makemeNA <- function(mydf, NAStrings, fixed = TRUE, overwrite = FALSE) {
  dfname <- deparse(substitute(mydf))
  if (!isTRUE(fixed)) {
    mydf <- data.frame(lapply(mydf, function(x) gsub(NAStrings, "", x)))
    NAStrings <- ""
  }
  mydf <- data.frame(lapply(mydf, function(x) type.convert(
    as.character(x), na.strings = NAStrings)))
  if (isTRUE(overwrite)) {
    assign(dfname, mydf, envir = .GlobalEnv)
  } 
  mydf
}