Is there a counterpart of the INMODEL option in PROC FMM?

[Season]

I am building finite mixture models and I wish to predict new observations using a model generated from PROC FMM. I found that this procedure has built-in capabilities of predicting the componentwise response as well as the mixing and posterior probabilities, which are all very useful. Therefore, I wish to directly carry a previously built model over to a dataset containing new observations and tell PROC FMM not to iterate in the modeling process. However, similar work can be easily done with the INMODEL option in PROC LOGISTIC in logistic regression modeling, but not in PROC FMM, as there is no such option in the latter procedure. What should I do then? Thanks!

[StatDave · Accepted solution]

See this general note on scoring. Try method 3.

[Season]

Thank you so much! I will try it.