statsmodels.discrete.discrete_model.DiscreteResults.bic

method

DiscreteResults.bic()

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