[SciPy-User] the skellam distribution

[josef.pktd at gmail.com]

On Mon, Nov 9, 2009 at 1:49 PM, Bruce Southey <bsouthey at gmail.com> wrote:
> On 11/09/2009 12:20 PM, Ernest Adrogué wrote:
>>   9/11/09 @ 17:01 (+0100), thus spake Johann Cohen-Tanugi:
>>
>>>    From what I understand of the initial statement from Ernest:
>>> "
>>> In case somebody is interested, or you want to include it
>>> in scipy. I used these specs here from the R package:
>>> cran.r-project.org/web/packages/skellam/skellam.pdf
>>> "
>>>
>>> he used the spec, as defined in this pdf, and did not look at the code
>>> itself. If my interpretation of the small preamble above is correct, I
>>> believe his implementation is not GPL-tainted, right?
>>>
>> Your interpretation of my preamble is correct.
>> I didn't look at the source code, I only read the PDF doc
>> where they explain in plain English how the pmf and cdf are
>> calculated.
>>
>>
> Okay,
> Then just provide a suitable reference outside the R package where
> someone can derive it independently or get more details. This also
> allows people to go back and check the implementation when things are
> not as expected. Perhaps the reference that Josef provided provides the
> same formulation?

If someone wants to check the relationship of chisquare and difference
of two independent poisson, but that's an implementation detail and
not really informative in a doc string.

On an Extension of the Connexion Between Poisson and χ2 Distributions
Author(s): N. L. Johnson
Source: Biometrika, Vol. 46, No. 3/4 (Dec., 1959), pp. 352-363

Josef

>
> Bruce
>
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>