I am using the loglin function of the base package to fit log-linear models. I am interested in obtaining the parameter values and their standard errors. Parameters are easily obtained, but I haven't found the way of obtaining their standad errors. Is this possible with the loglin function? If not, is there any other function to get them? Many thanks, -- Vicente Piorno Departamento de Ecolox?a e Biolox?a Animal EUT Forestal Campus Universitario 36005 Pontevedra SPAIN -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
It's a fundamental limitation of that method of fitting.
Most people fit log-linear models via surrogate Poisson models in glm or
via multinom {nnet}. There are worked examples of all the approaches in
MASS3/4.
On Wed, 23 Oct 2002, Vicente Piorno wrote:
> I am using the loglin function of the base package to fit log-linear
models.
> I am interested in obtaining the parameter values and their standard
errors.
> Parameters are easily obtained, but I haven't found the way of
obtaining
> their standad errors. Is this possible with the loglin function? If not, is
> there any other function to get them?
--
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Dear Vicente, At 06:31 PM 10/23/2002 +0200, Vicente Piorno wrote:>I am using the loglin function of the base package to fit log-linear models. >I am interested in obtaining the parameter values and their standard errors. >Parameters are easily obtained, but I haven't found the way of obtaining >their standad errors. Is this possible with the loglin function? If not, is >there any other function to get them? >Many thanks,The loglin function fits loglinear models by iterative proportional fitting, so standard errors for the parameter estimates are not produced. An alternative is to fit the model as a Poisson generalized linear model with the glm function, using the cell counts as the response; glm produces both estimated coefficients and standard errors, but be careful with the parametrization: it is conventional to parametrize loglinear models using sum-to-zero constraints (i.e., contr.sum). I hope that this helps, John ____________________________ John Fox Department of Sociology McMaster University email: jfox at mcmaster.ca web: http://www.socsci.mcmaster.ca/jfox ____________________________ -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- 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 _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._