Davis Vaughan

There does seem to be a little tension between this behavior and `step_select()`. Since `step_select()` requires you to specify the outcome to be able to keep it, to select just...

Alternatively, we could formalize what is generally required in a `step()` and then anything else that is required for a particular method could go in `$extras`. Like: ``` step

If this is mainly about step-window being slow, I agree that it is slow for wide window sizes as seen here. RcppRoll is used under the hood, and it fully...

Since the `...` are passed on to `arrange()` and I guess we want things like `desc(K)` to work, there isn't a _whole_ lot we can do here. `arrange()` is data-masking...

Wait on https://github.com/r-lib/pillar/issues/173 We could not use `vctrs::list_of` and make our own list class that holds `tbl_notes`, but I'd rather use vctrs

I tried that, and I think we need to do that, but the way it currently works with pillar is: - `pillar:::pillar_shaft.list` is called to generate the whole pillar column...

``` # We have: true positive rate = sensitivity(), recall() true negative rate = specificity() # We don't have: false positive rate = fall_out() false negative rate = miss_rate() ```...

@topepo in your example did you mean to say `pct = 1/2`? so we look for where it matches `.percent_tested`?

I think they are fine here because you only need workflows Imports/Suggests to test them and they are specific to workflows itself

To make it probably agnostic, this could be a metric that just counts the number of NA values present in `estimate`. `truth` might be required to work in a metric-set,...