Jan Sosulski

I will check today evening if I can fix the unit tests while still keeping their semantic logic the same as before. If not I will let you know.

I managed to fix the unit test for my case but broke several others by doing so. So either I will retry for a future release, or someone else can...

After some detachment of this issue, I think I was trying to misuse the droplog for my purposes. `epo.selection` should have all the info I need and I could subselect...

Nevermind, I cant figure out a way to make this work without re-structuring the drop_log / selection / events setup, because lets say you have first five epochs as target...

> For this specific use case could you make use of `write_events` to write a little companion `-eve.fif` or `-eve.txt` file alongside your epochs? Sure I can do that. I...

It is no urgent need for my use case, as I can just call `drop` after doing epoch subselection. Then the drop log stats work as intended. To fix the...

I dont think there is a straightforward/satisfying solution. As all bad epochs are already dropped, we can only slice-index on "good" epochs in our epochs object anyway. Hence slicing somehow...

After thinking about it, an acceptable solution could be to simply add 'IGNORED' to the drop log to the channel tuple when epochs were rejected, e.g. `( ('IGNORED', 'ch_1') ,...`...

When implementing this, there are two test failures in `epochs.py` caused by `equalize_event_counts` and direct comparisons with `('IGNORED',)`, but for now I cannot fix / understand the `equalize_event_counts` method in...

I pushed a draft PR #9768 . I have some issues with one unit test, which I think (maybe?) has a bug in it currently, but I am not sure...