[R] FFTW (Was: [R] Does R use the ATLAS linear algebra library?)

[Jason Turner]

> I'm
> also going to test out FFTW from http://www.fftw.org/; it looks good on
> paper, and if there are FFTs in R, it might be worth looking at.

FFTW is quite good.  Haven't got my benchmarks handy, but it is a 
noticible speedup, and it doesn't zero-pad.  The most obvious result
of not zero-padding is on the extreme frequencies (close to 0 or 0.5 
times the sampling frequency).  However, I believe the setup for 
comutation is different enough from what R uses internally that 
you'd have to set up another fft() and mvfft(), and load that in 
a library higher on the search path than "base".  Anyone done this, 
or something like it?

Cheers

Jason
-- 
Indigo Industrial Controls Ltd.
64-21-343-545
jasont at indigoindustrial.co.nz
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._