Hello, ? I believe that in R, it is not possible to analyze mixed effect-models when the distribucion is not gaussian (p.e. binomial or poisson), isn't? ? Somebody can suggest me alternative? ? thanks ? xavi ? -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Sure you can! See glmmPQL in package MASS, glmm in package glmmGibbs, glmm in one of Jim Lindsey's packages, .... There is even a discussion of this in MASS4 secion 10.4. On Mon, 21 Oct 2002, Xavi wrote:> I believe that in R, it is not possible to analyze mixed effect-models > when the distribucion is not gaussian (p.e. binomial or poisson), isn't?-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Mon, 21 Oct 2002, Xavi wrote:> Hello, > > I believe that in R, it is not possible to analyze mixed effect-models > when the distribucion is not gaussian (p.e. binomial or poisson), isn't?It depends on exactly what you mean. - Jim Lindsey's packages will fit (at least) random intercept models - For binomial or Poisson models with reasonably large means (perhaps 4 or so) the PQL approximation used by glmmPQL in the MASS package is pretty good.> Somebody can suggest me alternative?Again, it depends on why you want to fit mixed-effects models. You may be able to fit marginal models (GEE) instead. If you really want to fit mixed models with multiple random effects to binary data you probably need SAS PROC NLMIXED or a Bayesian solution (or HLM or MLWiN might be able to do it by now). -thomas -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Do I understand right, there is no multinom (nnet), with random effects available in R! Do I need to switch to SAS, PROC NLMIXED applying Agresti's example: http://stat2.uibk.ac.at/SMIJ/hartzel_abs.html Dominik> On Mon, 21 Oct 2002, Xavi wrote: > > > Hello, > > =A0 > > I believe that in R, it is not possible to analyze mixed effect-models > > when the distribucion is not gaussian (p.e. binomial or poisson), isn't> ? > > It depends on exactly what you mean. > > - Jim Lindsey's packages will fit (at least) random intercept models > > - For binomial or Poisson models with reasonably large means (perhaps 4 > or so) the PQL approximation used by glmmPQL in the MASS package is prett> y > good. > > > Somebody can suggest me alternative? > > Again, it depends on why you want to fit mixed-effects models. You may be > able to fit marginal models (GEE) instead. > > If you really want to fit mixed models with multiple random effects to > binary data you probably need SAS PROC NLMIXED or a Bayesian solution > (or HLM or MLWiN might be able to do it by now). > > -thomasDominik Grathwohl Biostatistician Nestl? Research Center PO Box 44, CH-1000 Lausanne 26 Phone: + 41 21 785 8034 Fax: + 41 21 785 8556 e-mail: dominik.grathwohl at rdls.nestle.com -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
> > Hello, > ? > I believe that in R, it is not possible to analyze mixed effect-models > when the distribucion is not gaussian (p.e. binomial or poisson), isn't?Several of the functions in my repeated library handle such data, available at www.luc.ac.be/~jlindsey/rcode.html Brian Riply also has a function that will do some glmms. Jim> ? > Somebody can suggest me alternative? > ? > thanks > ? > xavi > ? > > > > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html > Send "info", "help", or "[un]subscribe" > (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch > _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._ >-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._