mlwinfitMCMC object fitted values (propensity scores) storage

[tetroupploy]

I fitted a 2-level binomial model using mcmc. I want to apply predisposition score matching with calculated probability.
How can I convert mlwinfitMCMC fitted propensity scores (the predicted prob of Y for all obs) to numeric objects?

Base R commands are simple: p < model$fitted.values

[ChrisCharlton]

You should be able to do this with the overloads that R2MLwiN provides for the generic predict (https://stat.ethz.ch/R-manual/R-devel/l ... edict.html) function. To find more information on the syntax supported in R2MLwiN run:

Code: Select all

help("predict.mlwinfitIGLS")

or:

Code: Select all

help("predict.mlwinfitMCMC")

depending on which estimation method you are using.