[R-sig-ME] Random formula

[Luca Borger]

Hello,

given that you are fitting a gaussian response model and that your random 
effects are strictly nested (if I understand it correctly), nlme can be used 
easily (unless I get corrected by the experts on the list). You could fit 
them:

# m1, no random slope
response ~ climate_factor*pruning,
    random = list(provenance =~ 1, tree =~ 1),


# m2, random slope for pruning at provenance level
response ~ climate_factor*pruning,
    random = list(provenance =~ pruning, tree =~ 1),



HTH


Cheers,

Luca


----- Original Message ----- 
From: "De Smedt Sebastiaan" <Sebastiaan.DeSmedt at ua.ac.be>
To: <r-sig-mixed-models at r-project.org>
Sent: Monday, June 15, 2009 10:22 AM
Subject: [R-sig-ME] Random formula


> Hi,
>
> I measured leaf characteristics. The leaves are grouped in trees which 
> are, on their turn, grouped in provenances.
> I want to model those leaf characteristics in function of climate 
> variables (measured on provenance level) and pruning characteristics 
> (measured on tree level). I also want to see if the effect of pruning 
> differs between provenances (provenance-pruning interaction).
> The problem is that there cannot be an interaction between pruning and 
> tree, because pruning is measured on tree level.
>
> in lme4, I think I can specify this model as follows:
>
> response ~ climate_factor*pruning + (pruning|provenance) + 
> (1|provenance/tree)
>
> Or is there another way?
>
> Is it possible to define this model in the nlme library (I need a variance 
> structure, which doesn't exist in lme4)?
>
> Thanks a lot!
> Sebastiaan
>
> Sebastiaan De Smedt
> Department of Bioscience Engineering
> University of Antwerp
> Belgium
> Tel.: +32 (0)3 265 35 17
> Fax.: +32 (0)3 265 32 25
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>