fix vline when scrolling in epochs browser

[drammock]

this is item 4 from #8761

[jasmainak]

Thanks @drammock for giving a stab at this. When I tried to make a green vertical line by left clicking, this happened:

here's how I got to epochs:

import os
import mne

sample_data_folder = mne.datasets.sample.data_path()
sample_data_raw_file = os.path.join(sample_data_folder, 'MEG', 'sample',
                                    'sample_audvis_raw.fif')
raw = mne.io.read_raw_fif(sample_data_raw_file, verbose=False).crop(tmax=120)

events = mne.find_events(raw, stim_channel='STI 014')
event_dict = {'auditory/left': 1, 'auditory/right': 2, 'visual/left': 3,
              'visual/right': 4, 'face': 5, 'button': 32}
epochs = mne.Epochs(raw, events, tmin=-0.2, tmax=0.5, event_id=event_dict,
                    preload=True)

[drammock]

are you using the macos backend for MPL? ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐

…

On Friday, April 9, 2021 7:22 PM, Mainak Jas ***@***.***> wrote: Thanks ***@***.***(https://github.com/drammock) for giving a stab at this. When I tried to make a green vertical line by left clicking, this happened: [image](https://user-images.githubusercontent.com/15852194/114252009-35543f80-9971-11eb-8e08-869ca16056fc.png) here's how I got to epochs: import os import mne sample_data_folder = mne . datasets . sample . data_path () sample_data_raw_file = os . path . join ( sample_data_folder , 'MEG' , 'sample' , 'sample_audvis_raw.fif' ) raw = mne . io . read_raw_fif ( sample_data_raw_file , verbose = False ). crop ( tmax = 120 ) events = mne . find_events ( raw , stim_channel = 'STI 014' ) event_dict = { 'auditory/left' : 1 , 'auditory/right' : 2 , 'visual/left' : 3 , 'visual/right' : 4 , 'face' : 5 , 'button' : 32 } epochs = mne . Epochs ( raw , events , tmin = - 0.2 , tmax = 0.5 , event_id = event_dict , preload = True ) — You are receiving this because you were mentioned. Reply to this email directly, [view it on GitHub](#9276 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AAN2AU534B3ZPZDY346J4QLTH6K5ZANCNFSM42VXIJEQ).

[jasmainak]

so I had started using

$ ipython --matplotlib

what I thought was the recommendation but I just checked and it seems the recommendation is:

$ ipython --matplotlib=qt

and with this, it works!

[jasmainak]

Actually no, it doesn't work on that backend either. I spoke too soon. It depends where you click. Should I try another backend?

[agramfort]

I replicate the problem on my machine.

[larsoner]

I cannot replicate on my macOS machine, but I do get this:

(base) $ MPLBACKEND=MacOSX python -ui rep.py 
Traceback (most recent call last):
  File "/Users/larsoner/opt/miniconda3/lib/python3.8/site-packages/matplotlib/cbook/__init__.py", line 224, in process
    func(*args, **kwargs)
  File "/Users/larsoner/python/mne-python/mne/viz/_figure.py", line 563, in _resize
    old_width, old_height = self.mne.fig_size_px

And interaction with this backend is pretty bad. I recommend people use Qt5Agg on macOS, the experience will probably be better.

We should still fix this, but I wish the default backend on macOS was Qt5Agg so we didn't have to keep working around the MacOSX backend flaws in our code :(

[drammock]

I wish the default backend on macOS was Qt5Agg so we didn't have to keep working around the MacOSX backend flaws in our code :(

Another alternative is that one of you folks who has a mac laptop can do all the 2D viz work (or ship a mac laptop to me) :)