[R] User error in calling predict/model.frame

[Russell Pierce]

David,

Thanks for your feedback. However, I think if you look at the mean of
your output (which, as a minor point, doesn't provide the same data to
predict as expand.grid(X1=c(-1,0,1),X2=c(-1,0,1)), you'll see it is
very different from the mean of dat$out.  Like Ista's previous answer,
I think in your case the predict function acts on xxA and xxB (as if
renaming the newdata to dat[1:2] in Ista's case) and acts as if they
were in their original scale, that is, it doesn't treat the predictors
as Z scores.  As a result, the yhat values are different from what
should be expected.

-----------------------------------
Russell S. Pierce, M.A.
Visual Cognition Lab
Department of Psychology
University of California, Riverside
900 University Avenue
Riverside, CA 92521
Lab Phone: (951) 827-7399



On Sat, Jan 29, 2011 at 11:10 AM, David Winsemius
<dwinsemius at comcast.net> wrote:
> ms.res <- glm(out ~ scale(xxA)*scale(xxB), data=dat)
> newdata <- predict(rms.res, newdata=data.frame(xxA=c(-1,0,1),xxB=c(-1,0,1)))