Are there efficient approaches to compute p-values of samples drawn from a multivariate hypergeometric distribution? Specifically, how does the complexity grow as a function of the "urn size" and the number of different "colors" of balls within the "urn"?

mightbe suitable. – Bob bobson Jan 21 '11 at 4:32au faitwith p-value computation algorithms, would you provide some examples? That may help to anchor the question a bit. – David Roberts Jan 21 '11 at 4:53