[R-sig-ME] Formula Explanation in lme4 Vignette

Emmanuel Curis emmanuel.curis at parisdescartes.fr
Fri Nov 14 14:38:19 CET 2014


In lme4 formula syntax, random effects factors are on the right side
of the pipe (|). On the left side, you can have only 1 or other fixed
effects factors to include interactions between fixed effects and
random effects factors.

So this example does not violate your lecture, saying that a factor
can be with random or fixed effects, not both : Subject is the random
effect factor, Days the fixed effects, and Days | Subject introduces
an interaction between them --- that is, the Subject random effect is
different for each Days level.

Hope this helps,
Best regards,

On Fri, Nov 14, 2014 at 12:00:17AM +0000, Dario Strbenac wrote:
« Hello,
« In the software article vignette, it says that
« resp ~ FEexpr + (REexpr1|factor1)
« is the general format of mixed models. However, there's a sleep study example which has
« fm1 <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy)
« When I look at those two formulae, I see that Days matches both FEexpr and REexpr1. I thought that a factor could be a fixed effect or random effect, but not both. In lectures, I remember fixed and random effects as being exclusive, depending on if all the levels were measured, or just a subset of them.
« --------------------------------------
« Dario Strbenac
« PhD Student
« University of Sydney
« Camperdown NSW 2050
« Australia
« _______________________________________________
« R-sig-mixed-models at r-project.org mailing list
« https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

                                Emmanuel CURIS
                                emmanuel.curis at parisdescartes.fr

Page WWW: http://emmanuel.curis.online.fr/index.html

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