[R] Residuals of mixed effects model

will.eagle at gmx.net will.eagle at gmx.net
Thu Jul 29 16:07:27 CEST 2010


Dear all,

how do I get the residuals from a lme() output objects which are adjusted for fixed AND (!) random effects?

I tried residuals(), but it seems they just give me the residuals adjusted for the fixed effects of the regression model.

The model I use is:
lme.out <- lme(data=MyDataInLongFormat,fixed= outcome~1, random= ~ 1|individual, correlation=corSymm(form = ~time|individual))

Actually, I use only the intercept in the fixed part of the predictor, and I want to get residuals which are adjusted for the fixed part (intercept) and the random effect, ie to get rid of the correlatedness of individual measures across time. This way I want to get data where I can treat the measures per time point as independent groups. Makes sense?

Thanks in advance,

Will



More information about the R-help mailing list