Re: ROOT::Math::poisson_pdf(0,0)=NaN while TMath::PoissonI(0,0)=1

From: Lorenzo Moneta <Lorenzo.Moneta_at_cern.ch>
Date: Mon, 2 Aug 2010 12:12:02 +0200


Hello Corey,

 I agree that poisson_pdf(0,0) should return 1, although in theory the Poisson is defined only for value of the Poisson parameter greater than zero. I have modified now ROOT::Math::poisson_pdf in the ROOT trunk

 Thank you for your report

 Lorenzo

On Jul 28, 2010, at 1:55 PM, Corey Reed wrote:

> Hi,
>
> Just in case anyone (else) encounters this, I thought I would post it to this list.
>
> In Root 5.26.00/b,
>
> root [9] ROOT::Math::poisson_pdf(0,0)
> (double)nan
>
> root [11] TMath::PoissonI(0,0)
> (Double_t)1.00000000000000000e+00
>
> I personally would have expected the poisson_pdf to give 1 (the chance to get 0 events in the limit that the mean goes to 0).
>
> I don't know if this behavior is intended or not.
>
> - Corey
>
Received on Mon Aug 02 2010 - 12:12:06 CEST

This archive was generated by hypermail 2.2.0 : Mon Aug 02 2010 - 17:50:01 CEST