[R] apply lm.beta() to rlm object (robust regression)

robatok at gmx.de robatok at gmx.de
Tue Apr 19 00:56:20 CEST 2011


I'm trying to do a regression analysis (multiple linear regression) and have 
to deal with a slight heteroscedascitiy in my data.

I've read somewhere that it's possible to use the rlm (robust regression) 
out of the MASS package in such cases. Is it possible to apply the lm.beta 
method (from package QuantPsyc) to the returned rlm object and/or is there a 
way to calculate r-squared (or a "robust" equivalent) for each parameter?

Thanks a lot in advance.

More information about the R-help mailing list