[R] fit.contrast or some other for interactions and stratified

John Christie jc at or.psychology.dal.ca
Tue Apr 6 16:45:50 CEST 2004

I was trying to do a bunch of planned contrast but ran into a snag.  
fit.contrast won't work with an Error term in aov.  Ok, then I use lme. 
  Only, now it seems there is no way to enter the interaction terms 
(something like "x1:x2").  I couldn't find a way to use the interaction 
term in the help.  Is there an undocumented method or some other 
command for planned contrasts?

More information about the R-help mailing list