Thanks Gabe, and yes happy for you to submit the patch.
Some thoughts I've had in the interim:
1. Matt Dowle encouraged me to submit a patch as it could improve the
CRAN check farm timings since `ifelse` is presumably used a lot. I
thought the biggest benefit might come by improving the speed whe
Hugh,
(Note I speak for myself only and not for R-core) Thanks for looking into
this. I think it's great to have community members that are interested in
contributing to R and helping it continue to get better.
And I think, and my local experiments bear out, that using anyNA as a
fastpass conditi
Thanks Radford. I concur with all your points. I've attempted to address
the issues you raised through the github.io post. The new method appears
to be slower for test lengths < 100 and possibly longer lengths (not just <
10). Of course length(test) < 100 is very quick, so I simply added this to
t
> I propose a patch to ifelse that leverages anyNA(test) to achieve an
> improvement in performance. For a test vector of length 10, the change
> nearly halves the time taken and for a test of length 1 million, there
> is a tenfold increase in speed. Even for small vectors, the
> distributions of t
And the patch itself:
--- a/src/library/base/R/ifelse.R
+++ b/src/library/base/R/ifelse.R
@@ -18,12 +18,14 @@
ifelse <- function (test, yes, no)
{
+ attributes_of_test <- attributes(test)
+
if(is.atomic(test)) { # do not lose attributes
if (typeof(test) != "logical")
storage.mo
I propose a patch to ifelse that leverages anyNA(test) to achieve an
improvement in performance. For a test vector of length 10, the change
nearly halves the time taken and for a test of length 1 million, there
is a tenfold increase in speed. Even for small vectors, the
distributions of timings bet
