[Bug-gsl] wrong result in gsl_stats_variance for stride !=1 in gsl 1.3

[Christoph Marquardt]

Hello,

When using the gsl_stats_variance variance routine in GSL 1.3 on white noise 
data I saw different variances emerging when using different stride values 
(which shouldn't happen). 
I compared it with my own routine (text book algorithm) and my own routine had 
similar results for different strides, in contrast to gsl_stats_variance.

Unfortunately I only tested it with version 1.3, but I couldn't find a related 
bug in the archive...

Christoph
-- 
--------------------------------------------------------------------------
Christoph Marquardt

Institut fuer Optik, Information und Photonik
Max-Planck-Forschungsgruppe                     Tel.: +(49)9131-6877-121
Universitaet Erlangen-Nuernberg                 Fax: +(49)9131-6877-199
Guenther-Scharowsky-Str. 1 / Bau 24
D-91058 Erlangen              email: address@hidden