Hi Chris,
OK - stick with the RAM model, the h2 is so high you will run into
numerical issues otherwise. In the two-trait model you might want to add
in us(at.level(trait,1)):units into the random effects (make sure it is
not the last term in the random formula) in case log.dep has a h2
subst
Hi Jarrod,
I hadn't appreciated that the clustering of reproductive modes on the tree
might limit out ability to detect some of these relationships. This is in
fact a step in testing reproduction as an ordinal variable (egg-laying,
lecithotrophic live-bearing, and matrotrophic live-bearing) which f
Hi Jarrod,
Thanks very much for your fast reply. Egg-laying and live-bearing are
dispersed throughout the tree ( I have attached a PDF of a traitplot with
egg-laying and live-bearing on it; blue is egg-laying and red is
live-bearing), being universal in chimaeras and skates, and found in
several fa
Hi Chris,
I think MCMCglmm is probably giving you the right answer. There are huge
chunks of the phylogeny that are either egg-laying and live-bearing. The
non-phylogenetic model shows a strong relationship between reproductive
mode and depth, and that might be causal or it might just be becau
Hi Chris,
I think ngen in threshbayes is not the number of full iterations (i.e. a
full update of all parameters), but the number of full iterations
multiplied by the number of nodes (2n-1). With n=600 species this means
threshbayes has only really done about 8,000 iterations (i.e. about
1000
Hi All,
I am trying to look at the correlated evolution of traits using the
threshold model as implemented in phytools::threshBayes (Revell 2014) and
MCMCglmmRAM (Hadfield 2015). My understanding from Hadfield 2015 is that
the reduced animal models should yeild equivalent results, yet having run
bo