statsmodels.discrete.truncated_model.HurdleCountResults.bic

HurdleCountResults.bic

Bayesian information criterion. -2*llf + ln(nobs)*p where p is the number of regressors including the intercept.