Hi all,
I am trying to implement a parallelized version of my model using the reduce_sum function. I am wondering if it is somehow possible to still save the log-likelihood of each data point for model comparison later on. I don’t think there’s a way to save the log-likelihood values within the partial_sum function (where they are computed). Is it possible to have it return a vector containing the likelihood values of the subset of datapoints it’s doing the computation with and combine them into a vector containing the likelihoods of all datapoints in the model section? Or maybe there’s an alternative way that I am not thinking of.
Thanks!