[R-sig-ME] predict.lmer or lmer to lme conversion?

Adam D. I. Kramer adik-rhelp at ilovebacon.org
Sat Sep 13 21:25:17 CEST 2008


Hello,

 	I'm interested in making predictions from an lmer result:

> class(p1.lmer.additive)
[1] "mer"
attr(,"package")
[1] "lme4"

...naturally, I tried predict(p1.lmer.additive,newdata), but was told that
no acceptable method exists. The expected predict.lmer did not exist either.
I know that predict.lme exists for predictions based on the older lme()
output.

If I want predictions, should I refit my models with lme()? Is there a way
to convert an object of type "mer" to type "lme"? I'm not used to working
with S4 objects, so when names(p1.lmer.additive) is NULL, I get a bit
nervous.

Any suggestions appreciated.

--Adam




More information about the R-sig-mixed-models mailing list