Is it possible that a combination of independent variables are showing a perfect segregation in a binary logistic regression. As I am getting the warning: glm.fit: fitted probabilities numerically 0 or 1 occurred when using a number of variables in the model, but not individually when checked?


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.