Thank you for all this information. Another thing that I am trying to do is incorporate many features. Does it make sense to use a multivariate gaussian for this which has priors from another multivariate gaussian and LKJ Cholesky Covariance?
I implemented this an it take a 2 hours to run with just 4 individuals with 4 features while the above example would take approx 30 minutes. Is it likely I am messing up somewhere or these tend to take a long time?