If you have missing values for continuous variables that are used in
the model and you see the following:

* the binary or the multinomial distributions are specified

* there is an LSMEANS or an LSMESTIMATE statement that includes the AT
  or the ODDSRATIO option, or there is a MODEL statement that includes
  the ODDSRATIO option

* the observation that has a missing value in a continuous variable
  also has missing values in the response and no missing values in any
  class variables

the following error messages might appear:

   ERROR: Signal caught by CMP from PROC.
   ERROR: Invalid Operation.
   ERROR: Termination due to Floating Point Exception.

To prevent this problem from occurring, remove the observations that
have missing values in the analysis variables prior to running PROC
GLIMMIX.

Operating System and Release Information