[SciPy-user] Is this a bug in stats.geom.pmf?

[Stefan van der Walt]

On Mon, Jun 25, 2007 at 04:34:00PM -0600, Fernando Perez wrote:
> On 6/25/07, Stefan van der Walt <stefan at sun.ac.za> wrote:
> 
> > Which of these one calls "the" geometric distribution is a matter of
> > convention and convenience.
> > """
> >
> > So, do we simply pick one and stick with it?
> 
> I don't really care, but at least the code and the docstring should be
> consistent (and the docstring says that it uses the k>=1 formula).
> Also, the wikipedia page does list the (1-p)**(k-1)*p formula for the
> pmf, FWIW.
> 
> Those who actually use this stuff might want to chime in, but at least
> right now there is an inconsistency between code and docstring.

r3117 should fix this.  I used the (k-1)-version, since that is
what the original version was based on.

Cheers
Stéfan