statsmodels.discrete.discrete_model.MultinomialResults.bic

method

MultinomialResults.bic()[source]

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