replace_with_na throws error when replace list elements don't exist

[m-sostero]

replace_with_na throws an error when at least one of the replace list elements doesn't exist in the data, even if other elements do exist.

dat_ms <- tibble::tribble(
  ~x,  ~y,    ~z,
  1,   "A",   -100,
  3,   "N/A", -99,
  NA,  NA,    -98,
  -99, "E",   -101,
  -98, "F",   -1)

replace_with_na(
  dat_ms,
  replace = list(
    x = -99, # dat_ms$x exists
    w = -99  # dat_ms$w does not
  )
)

This gives

> Error in `[[<-.data.frame`(`*tmp*`, var, value = logical(0)) : 
  replacement has 0 rows, data has 5

I think the function return the data with all possible replacements, and throw an informative warning about replacement variables that do not exist in the data.

I will work on a PR for this.

[njtierney]

Ah, thank you for finding this, @m-sostero !

If you can work on a PR that would be fantastic!

[michael-dewar]

I don't know how to write a method, but an R wrapper like this should work:

safe_replace_with_na <- function(data, replace, ...){
	new_replace <- replace[names(replace) %in% names(data)]

	missing <- setdiff(names(replace), names(new_replace))
	if(length(missing) >0){
		warning(paste("Missing from data:", paste(missing, collapse = ", ")))
	}

	data %>% naniar::replace_with_na(new_replace, ...)
}

[njtierney]

Thanks for this, @michael-dewar! I've rolled this into replace_with_na and noted you in the NEWS file.