Try
weights = as.numeric(total)
BTW, there is a SIG (Special Interest Group) for lmer.
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
HTH,
--sundar
Chris O'Brien said the following on 8/14/2007 11:00
AM:> Dear R users,
>
> Prof. Ripley just corrected my understanding of the use of weights in glm,
> which I thought would allow me to correctly use lmer. However I'm
still
> having problems.
>
> My data takes the form of # of infected and uninfected individuals that
> were measured over time under different treatments. I'm using lmer to
> adjust for the repeated measures over time.
>
> In fitting the model:
>
> > model1=lmer(y~treatment+(time | ID),family=binomial,weights = total)
>
> where y = proportion of animals infected (number infected/total)
> total = number of infected + number uninfected
>
> this returns the error:
>
> Error in lmer(y ~ treatment + (time | ID), family = binomial, weights =
> total) :
> object `weights' of incorrect type
>
> Can anyone tell me what's wrong with the weights?
>
> cheers,
> Chris
>
>
>
>
>
> Chris O'Brien
> Sonoran Desert Research Station and
> School of Natural Resources
> Biological Sciences, room 125
> University of Arizona
> Tucson, AZ 85721
> work (520) 623-3720
> obrienc at email.arizona.edu
>
> ______________________________________________
> 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.