Correlated group-level effects with fixed correlation structure

I want to study how phylogenetic or trait dissimilarity affects covariance in species abundances. Specifically, I want to model species abundances in some plots multivariately, using group-level effects that correlate between species. In brms, such correlations are possible in the following way:

brm(mvbind(species_1,species_2,species_3) ~ (1|p|plot), family = “poisson”)

However, I would like to fix the correlation between random intercepts of species 1, 2, and 3 to some known matrix, so that the group-level effects could be thought to represent shared abundance responses due to phylogenetic relatedness or trait similarity. Is this possible with brms?

I know that the cov_ranef argument to brm can be used to specify a fixed correlation structure for the levels of a single group-level effect. It would be nice if this would be possible for correlation between group-level effects as well.

  • Operating System: Linux
  • brms Version: 2.10.0

@paul.buerkner might this be something you know about? Your BRMS-fu seems strong.

In future versions of brms, one will be able to fix parameters to constants which would presumably allow what was asked for here.