Hi all,

I have some difficulties in finding the max s(x). I'm using the code:

fit1=gam(y~treat+SEX+s(x,by=treat),data=dat,
     random=list(SUBJID=~1),correlation=corAR1(form=~1|SUBJID),
     optimizer=c("outer","newton"),method="GCV.Cp")

The bold part, s(x,by=treat) gives us 4 smoothing functions (4 levels in treatments). I would like to find out the maximum value of s(x) in each level.

Anybody could give me some hints??

  Please~  Thanks!!!