FIX: mne.io.read_raw_fil handling of bad channels
#12597
Conversation
|
I think I've attempted to be simultaneously too clever and not enough here. Pandas isn't a part of the minimal requirements so I should seek an alternative method to manipulate the |
georgeoneill
requested review from
larsoner,
drammock,
agramfort and
dengemann
as code owners
mne/io/fil/tests/test_fil.py
Outdated
| def _set_bads_tsv(chanfile, badchan): | ||
| """Update channels.tsv by setting target channel to bad.""" | ||
| data = [] | ||
| with open(chanfile) as f: |
There was a problem hiding this comment.
can you please explicitly pass an encoding here and below?
There was a problem hiding this comment.
I did in the orignal code, but ruff seems to be auto formatting them out...?
|
Now there's definitely a paper trail, I put rt/wt back into the code using the text editor on GitHub, and the CI style editor on here just took them back out. Thoughts @hoechenberger? |
|
The formatter removes those parameters because they're the defaults but I was talking about otherwise the encoding used will be platform-dependent |
Reference issue
Fixes #12594.
What does this implement/fix?
Changes the ordering that the
infostruct is populated inread_raw_filso that bad channels can be read straight from thechannels.tsvfile without raising an error.