On Aug 29, 2010, at 10:24 PM, David Winsemius wrote: > > On Aug 29, 2010, at 3:13 PM, moleps wrote: > >> >> glm(A~B+C+D+E+F,family = binomial(link = "logit"),data=tre,na.action=na.omit) >> Error in `contrasts<-`(`*tmp*`, value = "contr.treatment") : >> contrasts can be applied only to factors with 2 or more levels >> >> however, >> >> glm(A~B+C+D+E,family = binomial(link = "logit"),data=tre,na.action=na.omit) >> >> runs fine >> >> glm(A~B+C+D+F,family = binomial(link = "logit"),data=tre,na.action=na.omit) >> >> runs fine >> >> >> >> glm(A~E+F,family = binomial(link = "logit"),data=tre,na.action=na.omit) >> Error in `contrasts<-`(`*tmp*`, value = "contr.treatment") : >> contrasts can be applied only to factors with 2 or more levels >> >> Why is this? Could it be due to collinearity between the two? > > Perhaps, at least to the extent that the term "collinearity" is an > appropriate term for factor interactions. The obvious question at this point > is: What does: > > with( tre, table(E,F) ) # show?
Yes. Also notice the pattern of missing values. A likely scenario for the error would be to have E and F being something like "gender" and "menarche", where the latter is absent for boys. -- Peter Dalgaard Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Email: pd....@cbs.dk Priv: pda...@gmail.com ______________________________________________ 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.
