In our previous workflows, we include Goodness-of-Fit evaluation, which relies on some custom R code that runs diagnostics based on (parallelized) bootstrapping. Additionally, this GoF evaluation is then used to detect and iteratively remove outlies.
Is this something we want to include here as well? It would require including a bunch of our custom R, which we possibly have to rewrite to work with the current estimation API.
In our previous workflows, we include Goodness-of-Fit evaluation, which relies on some custom R code that runs diagnostics based on (parallelized) bootstrapping. Additionally, this GoF evaluation is then used to detect and iteratively remove outlies.
Is this something we want to include here as well? It would require including a bunch of our custom R, which we possibly have to rewrite to work with the current estimation API.