12-07-2011 02:19 PM
I am trying to separate two components (normal distributions) using a mixture model by PROC FMM. Assume I already know the two means of these two components. In PROC FMM, I can use RESTRICT to fix the means of these two components. However, I am also interested in the Bayesian analysis under PROC FMM. But RESTRICT does not work under Bayes statement in PROC FMM. Is there a way to fix these two means under Bayes Statement in PROC FMM? Thanks a lot.