I have performed a multinomial logistic regression using the multinom function in R. I have one categorical response variable with 3 levels, 3 continuous predictors and 3 categorical predictors (with 3, 4 and 5 levels). I thought about plotting the predicted probabilities in a facet style plot until I quickly realised that with this many variables it's not really possible to do that two-dimensionally.

Any suggestions for how to present predicted probabilities when you have multiple predictors (some continuous, some categorical), like this?

  • 2
    $\begingroup$ Can you provide some kind of reproducible example for people to work with? Are there any interactions in your model? What do you want to show & why, ie, what is the goal of the visualization? $\endgroup$ – gung Sep 30 '16 at 1:21

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.