Message-ID: <20130227144901.Horde.whrwJjp6yHZRLg7NGv9TyZA@webmail.csic.es>
Date: 2013-02-27T13:49:01Z
From: PALACIO BLASCO, SARA
Subject: lme4, error inserting nested structure of fixed factors in glmer
In-Reply-To: <20130227133221.GA14908@laboinfo-063.pharmacie.univ-paris5.fr>
Dear Emmanuel,
Thanks a lot for your response. I, however, do not think the full
model (i.e. "*" instead of "+") is the issue here, since I tried with
the simplified model before and it rendered the same error message as
the full factorial one...
Thanks in any case for your help!
Best regards,
Sara
> Dear Sara,
>
> ? - binary response variable: ?Dead? = bud survival, either dead (1)
> ? or alive (0)
> ? - Fixed factor: ?fTreatment?, numerical factor with 9 different levels
> ? - Fixed factor: ?fBud_type?, categorical factor with 3 levels
> ? - Fixed factor ?Species? (categorical) nested within ?fBud_type?,
> ? with 9 levels
> ? - Random factor ?fRep?, numerical, nested within ?Species?, with 24
> ? levels (i.e. coded sequentially to avoid confusion).
> ?
> ? The model I want to run is:
> ?
> ? M_bud_type1=glmer(Dead~fTreatment* fBud_type * fBud_type|Species +
> ? (1|fRep), family=binomial, data=species)
>
> Are you sure you want this model and not
>
> M_bud_type1=glmer(Dead~fTreatment + fBud_type + fBud_type|Species +
> (1|fRep), family=binomial, data=species)
>
> I guess you're problem comes from the a*b which stands for a+b+a:b,
> then coupling with more than two variables and interaction/nesting
> terms makes a not very clear formula.
>
> Try specifying only main effects and suited interactions, using either
> of the three syntaxes, and it my work better.
>
> Hope this helps,
>
> Best regards
>
> PS: this issue exists in all kinds of models, including
> lm/glm/lmer/lme...
>
>
> --
> Emmanuel CURIS
> emmanuel.curis at parisdescartes.fr
>
> Page WWW: http://emmanuel.curis.online.fr/index.html