Re: [Bug-gsl] cdf_poisson_P

[Rhys Ulerich]

This time with the list CCed...  Sorry I keep screwing up on the CC.

> I think there is an issue with function *gsl_ran_poisson_pdf* (unsigned int
> k, double mu):
>
> gsl_cdf_poisson_P(-1,5) returns 1.0000
>
> while
>
> gsl_cdf_poisson_P(1,5) returns 0.040428
>
> The second number is fine, but the first one should be either equal to the
> second one, or if the sign makes a difference, it should be zero.

If you turn on warnings, does the compiler shout at you for providing
a negative number (-1) as an unsigned argument (unsigned int)?

- Rhys