scipy.stats.rv_discrete. nnlf# rv_discrete.nnlf(theta, x)[source]# Negative loglikelihood function. Notes —– This is -sum(log pdf(x, theta), axis=0) where theta are the parameters (including loc and scale).