[R] lme or lmer for unbalance data

[Ben Bolker]

agent dunham <crosspide <at> hotmail.com> writes:

> I'd like to fix a mixed model. I have unbalance data, what should i use: 
> lme in nlme package , or lmer in lme4.
> 
> Thanks, user <at> host.com as user <at> host.com
> 

  More advanced mixed model questions belong on r-sig-mixed-models <at>
r-project.org , but the answer to this one is: either should be fine,
unless you want one of the features that only lme has (reported denominator
df/p values, R-side effects such as correlation and heteroscedasticity 
models) or one of the features that only lmer has (efficient fitting of
crossed random effects, GLMMs).

  Ben Bolker