There are several reasons, but the two most likely are 1. The model does not fit the data. 2. You are using variables that are linearly dependent. For example, if you include a binary indicator variable for "Males" and also include an indicator variable for "Females," the two variables are not independent and the covariance matrix is singular.   ... View more

Did you come up with a solution yet for CURE plot in Proc NLMIXED? Please share if you did. ... View more

You didn't ask a question, but  1. If you are asking why the model didn't converge, the most likely reason is that the model does not fit the data and/or you did not provide a good initial guess for the parameters. 2. If you are asking what to do about it, then    2a. Make a better initial guess for parameters    2b. Change to a different optimize, such as  TECH=NMSIMP or TECH=NRRIDG    2c.If you believe the model is almost converging, you can increase the number of iterations and the number of function evaluations that are permitted: MAXITER=500 MAXFUNC3000. ... View more

Here is the documentation for the NLMIXED procedure. The "Details" section discusses the two negative binomial parameterizations and shows how to use both in the procedure.  ... View more

I'm not a real user of proc nlmixed , I just went with its description. The negbin distribution is often difficult to fit with proc countreg or proc glimmix, so I would expect at least the same level of difficulty with proc nlmixed. You might ask for help about nlmixed on the SAS Statistical Procedures Community to find some expertise. Best of luck. ... View more