I am coming from R to Python for econometrics. In R, in case of fitting a logistic regression with glm, the output summary would include the result of test of the proposed model against the saturated model in the 'Residual Deviance' section. In Python, Statsmodels' glm output, however, I cannot see this particular test. Can somebody point me toward the right direction that does not involve fitting two models for every model I test?

Note that I am specifically asking for the following:

letting LL be log likelihood, where do I find 2(LL(Saturated_model) - LL(Proposed_model)) and the df = df(Saturated_model) - df(Proposed_model)?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.