Convergence warning when running proc hpgenselect

[ncmc]

I am running a generalized linear model using hpgenselect. Since I have too many parameters to estimate, I chose Techqniue=QUANEW to reduce the run time. While the model converged fine with the default NRRIDG, Quasi Newton did not converge and gave a warning message that the maximum number of iteration has met before convergence. My question is I still have estimates in the output log and ODS ParameterEstimate table. Does anybody know where these estimates are from? Are they from the previous iteration? Thank you in advance. 

[StatDave · Accepted solution]

They're from the last iteration.

[ncmc]

Thanks so much!