Jeffrey Pullin
Jeffrey Pullin
so that it can be efficiently extracted from models fit using grouped data and *lots* of items - like `voxels`.
Currently, we demand that all data be numeric - we should relax this to allow for, factors and character vectors etc. The key difficulty of this is how to present...
We could allow the user the specify the value of N and p in the prior. I haven't already done this because: * I'm trying to not create extra work...
For both `waic()` and `loo()`
Currently we implement `loo()` and (as of today) `waic()` to calculate model comparison metrics on the models fit using rater. Both functions, however, only support models fit using MCMC. This...
Currently, we implement initialization points that are basically the same as the prior. These were taken from one of Bob Carpenter's posts on the Stan forums I think - I...
https://mc-stan.org/users/documentation/case-studies/reduce_sum_tutorial.html