Using R I have fit some data to a model of the form y~A*B where A is a factor with levels A1, A2, ..., A5 and similarly with B.
I am using a different contrast than the default, so the results do not come out in terms of the original factor levels. I would like to take the estimates of the intercept and the contrasts and convert those back into estimates of the factor levels.
Is there a package or function for doing that? Also, perhaps more importantly, is there a standard name for doing that? If I just knew what it was called then I could probably track it down on my own, but I don’t even know the right terminology.