PROC GLIMMIX fits generalized linear mixed models. If the model does not contain random effects, PROC GLIMMIX fits generalized linear models by using the method of maximum likelihood. In the class of logistic regression models, PROC GLIMMIX can fit models to binary, binomial, ordinal, and nominal outcomes.