use the marginal distributions to extract type distributions within each cluster and observable characteristics

m2.mixt.stayers(sdata, model, ctrl)