On 5/22/2009 9:58 AM, Laura Bonnett wrote: > Dear All, > > I am attempting to use forward and/or backward selection to determine > the best model for the variables I have. Unfortunately, because I am > dealing with patients and every patient is receiving treatment I need > to force the variable for treatment into the model. Is there a way to > do this using R? (Additionally, the model is stratified by > randomisation period). I know that SAS can be used to do this but my > SAS coding is poor and consequently it would be easier for me to use > R, especially given the fractional polynomial transformations! > > Currently the model is as follows (without treatment). > > coxfita=coxph(Surv(rem.Remtime,rem.Rcens)~sind(nearma)+fsh(nearma)+fdr(nearma)+th1(nearma)+th2(nearma)+fp(cage)+fp(fint)+fp(tsb)+strata(rpa),data=nearma) > > > Thank you for your help, > > Laura
See the scope argument to stepAIC in the MASS package. You can specify a formula in the 'lower' component of scope which includes the treatment variable. That will force the treatment variable to remain in every model examined in the stepwise search. > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. -- Chuck Cleland, Ph.D. NDRI, Inc. (www.ndri.org) 71 West 23rd Street, 8th floor New York, NY 10010 tel: (212) 845-4495 (Tu, Th) tel: (732) 512-0171 (M, W, F) fax: (917) 438-0894 ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
