[R] leave-one-out cross validation in mixed effects logistic model (lme4)

Bert Gunter bgunter@4567 @end|ng |rom gm@||@com
Tue Sep 4 20:13:51 CEST 2018

Please post on the r-sig-mixed-models list, where you are more likely to
find the requisite expertise.

However, FWIW, I think the reviewer's request is complete nonsense (naïve
cross validation requires iid sampling). But the mixed models experts are
the authorities on such judgments (and may tell you that my opinion is
complete nonsense!).


Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

On Tue, Sep 4, 2018 at 10:16 AM Pedro Vaz <zasvaz using gmail.com> wrote:

>  Hello,
> So, I have this (simplified for better understanding) binomial mixed
> effects model [library (lme4)]
> Mymodel <- glmer(cross.01 ~ stream.01 + width.m + grass.per + (1|
> structure.id),
>   data = Mydata, family = binomial)
> stream is a factor with 2 levels; width.m is continuous; grass.per is a
> percentage
> Now, a reviewer is asking me to apply "a cross-validation procedure (i.e. a
> leave-one-out design coupled with predictive metrics as e.g. AUC) on this
> model"
> Does anyone have R-code to do this cross validation in my logistic mixed
> effects model? In the reviewer words: "the model should be evaluated also
> as for their predictive performance, not only for assumptions violation and
> for goodness-of-fit" (which I presented already in the reviewed paper
> draft)
> Many thanks in advance,
> pedro
>         [[alternative HTML version deleted]]
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.

	[[alternative HTML version deleted]]

More information about the R-help mailing list