Convert from string to double - “NAs introduced by coercion”

Question

I have a String List which looks like that:

> (eventWindow_120After_CDaxReturnList)
  [1] \"0,625955157\"  \"0,891329602\"  \"-0,933230106\"          


        

Answer 1

You have a problem with your decimal separators (R assumes . by default). The simplest solution is probably just to convert commas to periods (full stops)

testDat <- c("0,123","2,13","na")
as.double(gsub(",",".",testDat))

You will still get some warnings due to the "na" values in your data set; you could avoid the warnings by converting the "na" values separately, as follows:

NAvals <- testDat=="na"
res <- numeric(length(testDat))
res[NAvals] <- NA
res[!NAvals] <- as.double(gsub(",",".",testDat[!NAvals]))

There may also be some way to reset your locale (?Sys.setlocale) so that it works automatically, but I haven't been able to figure it out.

Here's another way to do it:

scan(text=testDat,dec=",",na.strings="na",quiet=TRUE)