statsmodels.discrete.discrete_model.LogitResults.bic¶ LogitResults.bic¶ Bayesian information criterion. -2*llf + ln(nobs)*p where p is the number of regressors including the intercept. Last update: Nov 14, 2024