# [SciPy-User] poisson distribution in scipy.stats

josef.pktd@gmai... josef.pktd@gmai...
Mon Nov 9 15:24:35 CST 2009

```On Mon, Nov 9, 2009 at 4:21 PM, Davide Cittaro
<davide.cittaro@ifom-ieo-campus.it> wrote:
>
> On Nov 9, 2009, at 10:03 PM, scipy-user-request@scipy.org wrote:
>
> Message: 5
> Date: Mon, 9 Nov 2009 15:00:29 -0600
> From: Robert Kern <robert.kern@gmail.com>
> Subject: Re: [SciPy-User] poisson distribution in scipy.stats
> To: SciPy Users List <scipy-user@scipy.org>
> Message-ID:
> <3d375d730911091300h6d9b4df6u4be1e6ae0811b12@mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> On Mon, Nov 9, 2009 at 14:58, Davide Cittaro
> <davide.cittaro@ifom-ieo-campus.it> wrote:
>
> Hi all,
>
> about the poisson generator... given l (expected) and k (found) I guess that
>
> the way to get the probability of k I have to do this:
>
> d = scipy.stats.poisson(l)
>
> p = pmf(k)
>
> Correct.
>
> which I found being the same of
>
> p = scipy.stats.poisson.pmf(l, k)
>
> Also correct.
>
>
> Although I've just plotted values and they are sligthly different... the
> second version does have only a max value at l, whereas the first has two
> maxima (l and l-1)
>
> I've here some code in which it is written:
>
> d = scipy.stats.poisson(l, k)

k is interpreted as location
same as stats.poisson(l, loc=k)

loc shifts the distribution, and returns a frozen distribution with
shifted support

Josef

>
> That one is completely wrong.
>
>
> Ok, I see... It looks like it shifts the pmf of k... but how does it works?
> I mean, how the discrete distribution constructor interprets this kind of
> declaration?
> d
>
> /*
> Davide Cittaro
> Cogentech - Consortium for Genomic Technologies
> 20139 Milano
> Italy
> tel.: +39(02)574303007
> e-mail: davide.cittaro@ifom-ieo-campus.it
> */
>
>
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>
```