[R] gam - Extraction of nonparametric component

[Michael A. Milligan]

Hello,

I am estimating a semiparametric partial linear model
using gam of the form

y=f1(x1)+f2(x2)+beta*X

where y is the dependent variable, f1(x1) and f2(x2)
are nonparametric functions of the independent
variables x1 and x2, respectively, and beta and X are
vectors of coefficients and independent variables. 
The R code is

EqGamAS	<-  gam(y ~ X+s(x1)+s(x2))

My question is, how can I extract the fitted values
of, say, f1(x1)?  Of course fitted(EqGamAS) returns
the fitted values of the entire regression function,
but is there a way to view only one component of the
nonparametric part of the estimation?  I have looked
through documentation and help archives and have not
found the answer.  I appreciate very much any help
anyone can give me.

Michael Milligan
Doctoral Candidate
University of New Mexico


      ____________________________________________________________________________________
Be a better friend, newshound, and