Re: [R] Weighted regression in rms/Hmisc

You should be able to pass a weights argument through fit.mult.impute to ols. I assume that weights are with respect to variables that you, for some reason, do not want to have as predictors. Otherwise you are already conditioning on the weighting factors and no weights are needed. Frank Mullah

[R] Weighted regression in rms/Hmisc

Hi all,  While using  the following package and code:    library(rms) library(Hmisc) test.trans<-aregImpute( ~ NearestWeekGestation + MaternalBMI + MomAge_Years +      WtGain + ethnicity , n.impute=10 ,data=data.1)    test.mod<-fit.mult.impute(BirthWeight_g ~ rcs(NearestWeekGes