For a linear regression example Y=X1+X2+e, I have only missing data in one of the covariates, say X1. What I did was to assign another linear model to X1, parallel to the model of Y in the model file. In this case, the DIC from dic.samples in rjags package is based on the joint likelihood of response Y and covariate X1? If I add another covariate to the model of Y, while the model for X1 remains the same, is the DIC comparable with the previous one?
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.