[R-sig-ME] MCMCglmm error message: MME singular: use a stronger prior

[Drager, Andrea Pilar]

Hi All,

I've received the MCMCglmm error message: Mixed model equations  
singular: use a (stronger) prior

In general, I'm curious what prior this refers to if one is specifying  
priors for both fixed and random effects?


More specific to my case:

My model is has a binary response, two continuous fixed effects and  
two random effects. I have a dataset of individual tree measurements  
for 23 species, and the issue that has caused this error is that one  
of the fixed effects, Zabund, is fixed per species, so that each  
individual of a species has the same value for this.

Error in MCMCglmm(tally ~ Zabund + Zdbh + Znnd, random = ~phylo +  
code, family = "categorical",  :
   Mixed model equations singular: use a (stronger) prior

priorI = list(R = list(V = 1, nu = 0, fix = 1), G = list(G1 = list(V =  
1,nu = 0),
                                                          G2=list(V=1,nu=0)))

Thanks for any insight,


Andrea Pilar Drager
PhD. student
Ecology and Evolutionary Biology, Rice University