####
Correlated Data: The GENMOD and GLIMMIX Procedures

When a generalized linear model is formed by using distributions other than the binary, binomial, or multinomial distribution,
you can use the GENMOD and GLIMMIX procedures for parameter estimation and inference.

Both PROC GENMOD and PROC GLIMMIX can accommodate correlated observations, but they use different techniques. PROC GENMOD
fits correlated data models by using generalized estimating equations that rely on a first- and second-moment specification for the response data and a working
correlation assumption. With PROC GLIMMIX, you can model correlations between the observations either by specifying random
effects in the conditional distribution that induce a marginal correlation structure or by direct modeling of the marginal
dependence. PROC GLIMMIX uses likelihood-based techniques to estimate parameters.

PROC GENMOD supports a Bayesian analysis through its BAYES statement.

With PROC GLIMMIX, you can vary the distribution or link function from one observation to the next.

To fit a generalized linear model by using a distribution that is not available in the GENMOD or GLIMMIX procedure, you can
use PROC NLMIXED and use SAS programming statements to code the log-likelihood function of an observation.