kernel smoothers for gam( )

["Mark Stevenson" <M.Stevenson@massey.ac.nz>]

Hello,

Chambers and Hastie (Statistical Models in S) state that user-defined
smoothers can be used in fitting additive models in gam() (page 293).

I'm wanting to incorporate a kernel smoother (rather than the loess and
smoothing spline functions already 'packaged' with the gam function). Can
anyone out there point me in the right direction for a kernel function that
I can 'bolt' straight into gam()?

Will summarise to the list.

Regards,

Mark S

*******************************************
Mark Stevenson
EpiCentre, IVABS
Massey University
Private Bag 11-222 Palmerston North
NEW ZEALAND

Ph: +64 6 350 6149
Fx: +64 6 350 5716
M.Stevenson@massey.ac.nz
http://www.dairywin.co.nz/
*******************************************