First of all, the documentation is not very clear. The input arguments are not explained so it took me sometime to find out how it is parameterized. CumulativeDistribution calculates P(X < x) not P(X <= x).
var hyper = new Hypergeometric(10000, 2, 9800); // 2 successes, 99998 failures, 9800 draws
double p = hyper.CumulativeDistribution(0.1); // P(X <= 0)
p is NaN for 9800 and many other values.
Comments: Thanks for reporting this! I can confirm all 3 issues.
var hyper = new Hypergeometric(10000, 2, 9800); // 2 successes, 99998 failures, 9800 draws
double p = hyper.CumulativeDistribution(0.1); // P(X <= 0)
p is NaN for 9800 and many other values.
Comments: Thanks for reporting this! I can confirm all 3 issues.