Re: [Bug-gsl] Error in documentation of real FFT

[Brian Gough]

At Mon, 12 May 2008 15:25:12 +0200,
Cristian Toraci wrote:
> The documantation for gsl_fft_real_radix2_transform says that the
> output can be sorted using gsl_fft_halfcomplex_unpack. The
> documantation of gsl_fft_halfcomplex_unpack says it can also be used
> to sort the output of gsl_fft_real_transform, but the output of
> gsl_fft_real_radix2_transform and gsl_fft_real_transform don't have
> the same storage. Actually I verified than
> gsl_fft_real_radix2_transform + gsl_fft_halfcomplex_unpack doesn't
> sort the data in the right way: the result is different from what
> you get converting the data to a complex array and transforming it
> with gsl_fft_complex_radix2_forward, but the result is correct if
> the output of gsl_fft_real_radix2_transform is sorted by hand.

Thanks for your email.  The documentation should have referred to the
separate function gsl_fft_halfcomplex_radix2_unpack -- I will correct
that.

-- 
Brian Gough

GNU Scientific Library -
http://www.gnu.org/software/gsl/