Dear stan
community, I am facing the following programming problem writing my own log probability mass function for the multivariate Fisher & Wallenius’ non central hypergeometric distributions.
Given two positive integers n and c>1, I would like to build a matrix of integers with c columns and
\frac{(n+c-1)!}{n!(c-1)!} rows.
To calculate the factorials i use the tgamma()
function, so that \Gamma(n+1) = n!.
However tgamma
returns real values and I cannot pass real values a s dimension of a matrix.
Any workaround?
Thanks in advance.