partial scoring

Quartz | Level 8

Hello,

I would like to score a dataset with only part of a model estimated through proc genmod.

In other terms, I have fitted the following regression model with many covariates (mostly categorical dealt with the class statement)

y=b0+b1*x+b2*z+epsilon

where x and z are vectors of covariates (a number of variables with many levels each).

I would like to get, on another dataset, the predicted values: y=b0+b2*z.

Is it possible with proc plm or proc score?

Any advice is appreciated.

Thank you very much in advance