disabling mapply automatically converting Dates to numeric

Question

I've noticed that when you do this:

mapply(function(x) { x }, c(as.Date('2014-1-1'), as.Date('2014-2-2')))

R automatically converts your vector of Dates into a vector of numbers. Is there a way to disable this behavior?

I know that you can wrap the result in as.Date(..., origin='1970-1-1'), but I can only imagine there has to be a better solution here.

side note : You should maybe rewrite your code example , this is not the right way to use `mapply`. — agstudy, Dec 17 '14 at 23:14
I mean that the usual way to use `mapply` is `mapply(function(x,y),X,Y)` for example `mapply(rep, 1:4, 4:1)`. You are using it as `sapply/lapply`. As I already mentioned this only a side note. — agstudy, Dec 17 '14 at 23:34

Rich Scriven · Accepted Answer · 2014-12-17T22:57:01.487

6

This has to do with the way mapply simplifies its result through simplify2array.

x <- list(as.Date('2014-1-1'), as.Date('2014-2-2'))
simplify2array(x, higher = FALSE)
# [1] 16071 16103

You can turn off the simplification and then reduce the list manually.

do.call(c, mapply(I, x, SIMPLIFY = FALSE))
# [1] "2014-01-01" "2014-02-02"

Or you can use Map along with Reduce (or do.call)

Reduce(c, Map(I, x))
# [1] "2014-01-01" "2014-02-02"

Map is basically mapply(..., SIMPLIFY = FALSE) and I use I in place of function(x) { x } because it just returns its input as-is.

edited Dec 17 '14 at 22:57

answered Dec 17 '14 at 22:39

Rich Scriven

Using `I` instead of `(` makes more sense to me. – Dason Dec 17 '14 at 22:52
Oh yeah, that works too. Didn't think about that. Made the edit – Rich Scriven Dec 17 '14 at 22:53
I like the Reduce(c, Map(I, x)) approach the best of everything discussed. Thanks! – chuck taylor Dec 17 '14 at 23:34
@chucktaylor FYI , `Reduce` is generally slower than `do.call`. – agstudy Dec 17 '14 at 23:36

1 Answers1