| To: | Michel Friesenhahn <michel.friesenhahn.b@bayer.com>, s-news@lists.biostat.wustl.edu |
|---|---|
| Subject: | Re: Cubic Smoothing Splines |
| From: | impromptu@nifty.com |
| Date: | Thu, 19 Feb 2004 13:15:35 +0900 |
| In-reply-to: | <OF56143A3F.90EB47D4-ON88256E3F.000B41CD-88256E3F.000BD97A@bayer.com> |
| References: | <OF56143A3F.90EB47D4-ON88256E3F.000B41CD-88256E3F.000BD97A@bayer.com> |
Hi! Mike and S-Plus users
>I'm fitting cubic smoothing splines with smooth.spline(). Does anybody know
>how
>to get the cubic polynomial coefficients for each of the knots? I've used
>looked at the $fit$coef component for the output of predict() but can't make
>sense of it.
I hope that the object below will help:
function()
{
x1 <- seq(from = 1, to = 20, length = 20)
y1 <- sin(x1 * 0.1)
fit.sp <- smooth.spline(x1, y1, spar = 1e-010,
all.knot = T)
print(names(fit.sp$fit))
print(fit.sp$fit$knot)
print(fit.sp$fit$coef)
}
***** Kunio Takezawa, Ph.D. (takezawa@affrc.go.jp) *****
***** <http://cse.naro.affrc.go.jp/takezawa/patent-c.html> *****
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Cubic Smoothing Splines, Michel Friesenhahn |
|---|---|
| Next by Date: | Re: NLME loop termination, Pete Meyer |
| Previous by Thread: | Cubic Smoothing Splines, Michel Friesenhahn |
| Next by Thread: | lme and aov for for fractional design with a random effect, Olivier Renaud |
| Indexes: | [Date] [Thread] [Top] [All Lists] |