mixed radix fft (radix 2 and radix 5)

I am looking for a mixed radix 2 and 5 fft algorithm in C. I have
download some algorithm from www.fftw.com but don't know how to
install it(I am using windows XP not unix). Basically I have to do a
10, 100, 1000,... fft by using mixed radix(2,5)algorithm. Does anyone
can give me some information about the algorithm? Where can I
download?

Thanks!

Jenny

qiuping18@hotmail.com (Jenny)...
> I am looking for a mixed radix 2 and 5 fft algorithm in C. I have
> download some algorithm from www.fftw.com but don't know how to
> install it(I am using windows XP not unix). Basically I have to do a
> 10, 100, 1000,... fft by using mixed radix(2,5)algorithm. Does anyone
> can give me some information about the algorithm? Where can I
> download?

(It sounds like you are looking for an implementation, not an
algorithm, but I digress.  Also, fftw.com is an investment firm; you
mean fftw.org, I think.)

We don't use Windows ourselves, but you can download Windows
installation instructions, as well as precompiled binaries, from our
web page (www.fftw.org/install/windows.html).

Alternatively, there are a few other C codes that handle mixed
factorizations and are available online.  See our benchmark
(www.fftw.org/speed) for the different codes that we compare for
non-powers-of-two (links at http://www.fftw.org/benchfft/ffts.html).

Cordially,
Steven G. Johnson

qiuping18@hotmail.com (Jenny) wrote in message news:<a4d3198d.0309172330.3cb63028@posting.google.com>...
> I am looking for a mixed radix 2 and 5 fft algorithm in C. I have
> download some algorithm from www.fftw.com but don't know how to
> install it(I am using windows XP not unix). Basically I have to do a
> 10, 100, 1000,... fft by using mixed radix(2,5)algorithm. Does anyone
> can give me some information about the algorithm? Where can I
> download?
> 
> Thanks!
> 
> Jenny

An algorithm of the type you are looking for is in netlib.
Try at 
http://gams.nist.gov 
or google for "Clive Temperton" who did the work on which the radix
2,3,5 algorithms I know are based. If you cannot find anything I could
mail a variant of the netlib code (my mail adress is "spammable").
(note: in part f2c converted code, not extremely legible but works -
minor modifications to the type declarations might be necessary for
Visual C++ because I use it with gcc on windows)

Sorry, I don't like/use fftw and cannot assist with the
implementation.

Robert

Robert Malek wrote:
> 
> Sorry, I don't like/use fftw and cannot assist with the
> implementation.

I'm curious. What is your objection to FFTW?

Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo  nospam@mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
Microsoft is not the answer.
Microsoft is the question. NO is the answer.

yeren@gmx.de (Robert Malek) wrote...
> Try at http://gams.nist.gov 
> or google for "Clive Temperton" who did the work on which the radix
> 2,3,5 algorithms I know are based.

Actually, the oldest mixed-radix algorithm is by Gauss (and Cooley and
Tukey also described the mixed-radix case, and even thought that the
optimal radix was 3!).  The earliest mixed-radix source code I can
think of offhand is by Singleton in 1968 (code is still available at
netlib).

Of course, you may be referring to the "GPFA" code by Temperton, which
uses a mixture of the Cooley-Tukey algorithm and the Prime-Factor
(Good-Thomas) algorithm if I recall correctly, rather than "straight"
mixed-radix Cooley-Tukey.  (Prime-Factor only works for factors that
are relatively prime, e.g. to separate all of the powers of two from
all of the powers of 3.)  Usually, I find FFTPACK to be faster than
the GPFA code (see our benchmarks), though.  (All of these codes are
in Fortran, which is why I didn't mention them in my earlier posting. 
Of course, if you are willing to use f2c...)

Steven

> I'm curious. What is your objection to FFTW?
To tell the truth: I think it is too complex and big a package for my
purposes. With a piece of sourcecode I don't need to "install" some
package for different platforms.

Robert