Anthony R. Ives

This might or might not be a technical issue, but it is puzzling. Using bayes = T in a poisson model with (1|site) and (1|plot) where plots are contained within...

need for logLik for family=Poisson

3

For bayes = F, family=Poisson doesn't produce logLik, although family=binomial does. I thought the code was there for the logLik, but I might need to write it.

predicted values with INLA

4

Russell, While updating rr2, I found that the bayes=T version of pglmm returns the "nearest_node" predicted values even when the "tip_rm" option is used. Is there a structural reason for...

Unlike other distributions, we do not automatically include a (1|obs) term for the distributional component of the zero-inflated models. I'm working on an example in which this matters. It can...

communityPGLMM.plot.re()

4

I couldn't get this to work. There were a variety of error messages, such as Error in as(as.matrix(Y.mat), "denseMatrix")[, cov_ranef_update[[sp.var]]$tip.label] : invalid or not-yet-implemented 'Matrix' subsetting It would be nice...

communityPGLMM.bayes

21

Russell, I've just tried communityPGLMM.bayes with a binomial distribution in the cbind(success, failure) format, and I get the error Error in INLA::inla(as.formula(inla_formula), data = data, verbose = verbose, : dim(y...orig)[2]...

binaryPGLMM

2

I figure we should update binaryPGLMM to do family=binomial and poisson. Should we change the name? This has the advantage of differentiating it from binaryPGLMM in ape, which we've talked...

structuring covariance matrices in communityPGLMM

7

Just some thoughts: 1. I've been working on a bipartite problem, and the simple format of communityPGLMM with repulsion=T sets both site and sp to the repulsion matrix. To look...

communityPGLMM.plot.random.effects doesn't work well

2

When you have a lot of matrices, all the plotting region is filled by the legend bar. I suggest going back to using image() [since I never found a way...