Giuseppe <neox86 <at> gmail.com> writes:
>
> Dennis,
>
> Yes, Reduce works, but unfortunately it does not speed up the
> evaluation time (actually, it makes it a little bit slower).
>
> Thank you anyway.
I have an alternate solution that I would have expected to
be faster but is actually much slower for this tiny example.
I even tried testing it on larger matrices: I thought that there
was an outside possibility that the startup takes longer but that it
scales better with system size
... ? (I generally think apply/sweep should be faster than Reduce/lapply?)
Oh well.
m <- array(1:27, c(3, 3, 3))
xx <- 1:3
system.time(replicate(1e5,
Reduce('+', lapply(xx, function(k) xx[k] * m[, , k]))))
## 5 seconds
system.time(replicate(1e5,
apply(sweep(m,3,xx,"*"),c(1,2),sum)))
## 22 seconds
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
