Thomas L Jones wrote:> I am working with a non-parametic smoothing operation using a
> Generalized Additive Model. It is a bivariate data set. I know how to
> do the smooth, and out comes a nice smooth curve.
>
> Now I want to find the value of the smoothed curve for several values
> of x (the abscissa). This can be done (please correct me if I am
> wrong) by using the predict.gam function. You feed the predict.gam
> function a data frame, telling it what you want.
>
> Let's start with the predict.gam function. Are you supposed to be able
> to look up how to use it? E.g., what goes into the various columns of
> the data frame? I do have a working function call for predict. It
> says:
>
> pred_out <- predict (mod, data.frame (x = x), type =
"response")
>
> (mod is the model)
>
> Now, if you tell me that x = x, I will believe you. But what is meant
> by "data.frame (x = x)", I know not.
>
> Or would it better to call the class, names, and str functions, using
> some well chosen objects?
Does this help?
library(gam)
data(kyphosis)
kyphosis.gam <- gam(Kyphosis ~ s(Age,4) + Number, family = binomial,
data=kyphosis)
predict(kyphosis.gam, data.frame(Age = c(10,100,200), Number = 3),
type="response")
1 2 3
0.019098701 0.248043366 0.005807921
The data frame should contain the predictor variables in the model and
the values for the variables should be those at which you want a prediction.
> Tom Jones
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
--
Chuck Cleland, Ph.D.
NDRI, Inc.
71 West 23rd Street, 8th floor
New York, NY 10010
tel: (212) 845-4495 (Tu, Th)
tel: (732) 512-0171 (M, W, F)
fax: (917) 438-0894