Open jrosell opened 6 days ago
Could you say a little bit more about why it is that you'd need to extract the splits from the tune_results
rather than just reusing the splits you have already?
Note to self: FWIW, we did find a use for a similar helper in stacks:::.set_splits()
.
Well. In my pipelines I usually have one process for fitting resamples & tuning and sometimes I only save the tune_resamples object and not the rset... But, then "ups" I need the rset too because I want to check something and I didnt save it. {tailor} could increase the probability of this issue.
Furthermore, I want to try AutoGuon inference approach and this function could help.
Feature suggestion
Now that we have the new {tailor} package for post-processing in titydmodels, I find myself in the need to reuse the splits from tune_results as a resampling object.
I believe this new extract_resamples function (or whatever name you prefer) could improve the interactive usage of tidymodels.
Here a minimal reproducible example to demonstrate its use:
This implementation seems to give identical results for my vfold_cv example, but I guess other rset type of objects should be tested.