inmoose.utils.rnbinom

inmoose.utils.rnbinom(n, size, mu, seed=None)

mimic R rnbinom function, to draw samples from a Negative Binomial distribution.

The (\(size\), \(p\)) parameterization used in R is the same as in scipy.stats: \(p = 1 / (1 + \mu/size) = size / (size + \mu)\).

Parameters:

• n (int or tuple of ints) – shape of the output. If n = (n1, n2, …, np) then n1*n2*…*np random samples are drawn.

• size (float or array-like) – size parameter of the Negative Binomial distribution. all values must be positive

• mu (float or array-like) – mean parameter of the Negative Binomial distribution all values must be positive

• seed (int, optional) – pass a seed to the underlying RNG. If None, then the RNG is seeded using unpredictable entropy from the system. See the documentation of scipy.stats about RNG seeding for more details.