[S] Point-Wise Confidence Intervals for a GAM

Steven Abrahams (swa@ms.com)
Mon, 26 Oct 1998 18:42:18 -0500


This is my first time e-mailing to this address, but it seems a
promising source of help on a very specific Splus question. In advance,
thank you for any advice you can provide.

The Splus function predict.gam allows the user to specify se=T to get
pointwise standard error bands around a fitted value. What elements of
the gam object does this function use to estimate the standard errors,
and with what formula. I want to be able to reconstruct the
calculation. I have skimmed through Hastie and Tibshirani's General
Additive Models for a specific answer, but so far I'm coming up empty.
Any suggestions?

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news