[R-sig-ME] Speed estimation for lmer?
HDoran at air.org
Thu Sep 11 23:17:24 CEST 2008
Choosing the identity link in glmer is the same as a linear model. You
can test for speed using system.time(), but there may be internal code
reasons why one would prefer glmer over lmer, that I don't know.
> -----Original Message-----
> From: r-sig-mixed-models-bounces at r-project.org
> [mailto:r-sig-mixed-models-bounces at r-project.org] On Behalf
> Of Daniel Ezra Johnson
> Sent: Thursday, September 11, 2008 5:07 PM
> To: r-sig-mixed-models at r-project.org
> Subject: Re: [R-sig-ME] Speed estimation for lmer?
> >> Roughly speaking: lmer and lmer2 aren't (I think) different any
> >> more, they were different branches of the same software.
> They should
> >> both be much faster than lme. glmer (from lme4) and glmmPQL (from
> >> nlme) should not be necessary unless you have binomial,
> Poisson, etc.
> >> data rather than normally distributed responses.
> > Ben is correct. There are functions lmer and lmer2 in the current
> > lme4 package but lmer2 is just a stub that turns around and calls
> > lmer. If your response y is on a continuous scale then it
> should be
> > lmer that you use.
> so if my response is on a continuous scale, is there any
> difference between
> lmer(y~fixed+(1|random)) and
> glmer(y~fixed+(1|random),family="gaussian") ?
> the output is the same but since reading the above i'm
> wondering if 'plain' lmer is written to run faster?
> R-sig-mixed-models at r-project.org mailing list
More information about the R-sig-mixed-models