tune
tune copied to clipboard
pass only `labels(split)` to internal functions that just take `split` for its `labels()`
In tune_grid_loop_iter(), we pass split to several error handling functions that only use the split to generate informative error messages, i.e. with labels(split). We should instead just write:
split_label <- labels(split)
at the top of the function and rewrite those helpers to just take the label so that it's clearer why that's being passed around. This will be especially helpful once a post-processor is plumbed through that further splits split to properly evaluate the calibrator.