Sampler initialization error with a model containing an LKJCholeskyCov distribution

Thank you again @Simon for taking the time to understand where the error in this code sample came from and for the solution you provided!

This is also what I ended up doing, i.e. using the names of the un-transformed parameters in the initialization dictionary supplied to the sampler, even though the documentation suggests otherwise.

Of course the most important thing is that it works, but it is still a bug I think, albeit a minor one.

1 Like