[R-sig-ME] Question of Brasil

Ben Bolker bbolker at gmail.com
Mon Mar 24 02:18:53 CET 2014


On 14-03-23 08:34 PM, Alejandra Tapia wrote:
> Dear Ben,
> 
> I have two question:
> 
> 1) In lme4 package  when we use the function glmer with Laplace
> Aproximation (nAGQ=1) or Adaptative Gaussian Quadrature (nAGQ=5), the
> random effects by which method are calculated?
> 
> 2) For to glmer or command resid() using the same method that 1) for
> calculate or random effects?
> 
> Thank's
> 
> Greetings
> -- 
> Alejandra
> IME-USP, Brasil

  I'm taking the liberty of cc'ing this to the r-sig-mixed-models list,
because I really can't answer private lme4 questions, no matter how much
I would like to -- I would get no time to do anything else.

  I'm afraid, in addition, that I don't understand your questions very
well.  (1) Standard references on GLMMs such as Bolker et al 2008
explain what Laplace approximation and Gaussian quadrature are: if you
need more specific information, you'll have to rephrase your question
more precisely. (2) I really don't understand this at all. Can you
rephrase or give an example?

  sincerely
    Ben Bolker
>



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