Summary of feature requests from this discussion:

• epoching (fixed in #45)
• epoch selection (by condition/event)
• epochs concatenation (will be fixed in #131)
• epochs averaging (evoked)
• ERP plotting - this should work once epochs averaging is implemented

It would be better if MNELAB didn't use a .qrc file, but included icons directly. One of the reasons is that PyQt5 and PySide2 use slightly different command line tools (pyrcc5 vs. pyside2-rcc) and produce files which include PyQt5.QtCore or PySide2.QtCore. This is problematic when using QtPy, because these automatically generated files would need to be edited manually. Therefore, getting rid of the resource system completely would probably be the best solution in the long run.

From the associated PR #4:

Currently, the channel picker dialog does a good job of allowing the user to
choose bad channels, but has a few shortcomings:

(1) It's difficult to see, at a glance, which channels have already been marked
as bad.

(2) When the user begins clicking on new channels to mark as bad, any
previously marked channels become unselected. Thus, making sure you've selected
everything you want requires careful scrolling.

This change introduces a few new files (including a .ui file from QtDesigner and generated .py from the pyuic5 tool) and @cbrnr might have preferences about how this sort of thing is organized/implemented. Since this might entail multiple PR's, an issue might be a better place for discussion.

Please feel free to review my code and/or let me know your thoughts and ideas.

Using the default command pip install mnelab on Windows I get:

Installing collected packages: PyQt5, mnelab
Could not install packages due to an EnvironmentError: [WinError 5] Access is denied: 'C:\\Users\\mmagn\\Continuum\\anaconda3\\Lib\\site-packages\\PyQt5\\Qt.pyd'
Consider using the `--user` option or check the permissions.

strangely, I get it even when I run it in command prompt with administrator privileges. When I run using pip install mnelab --user the script the mnelab.exe is put in a separate directory and does not do anything. When I then first activate my base anaconda env (it is my default python so I don't know why it makes any difference) and then do pip install mnelab - I get everything installed in one directory but mnelab.exe still does not work.
When I try to import mnelab in python I get:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "C:\Users\mmagn\Continuum\anaconda3\lib\site-packages\mnelab\__init__.py", line 1, in <module>
    from .mainwindow import MainWindow
  File "C:\Users\mmagn\Continuum\anaconda3\lib\site-packages\mnelab\mainwindow.py", line 7, in <module>
    from PyQt5.QtCore import (pyqtSlot, QStringListModel, QModelIndex, QSettings,
ModuleNotFoundError: No module named 'PyQt5.sip'

I might have been unlucky - I can test tomorrow on other Windows computers.

Even though MNELAB uses SVG icons, they must have been converted to low-res bitmaps somewhere along the way (in the Qt resource system?).

Currently, the button is grayed out until one sets a montage via the Edit menu. Maybe it should always be active, but present the user with the Set montage dialog box if no montage has been set.

Consider providing a visual programming interface for building on/offline EEG analysis flows.

User story

As an end user
I would appreciate a visual programming UI
so I can easily build pipelines for my EEG data analysis

Goal

Allow MNE-Lab users to create EEG analysis flows by dragging, dropping, connecting, and configuring modules via a visual programming interface.

Examples

PyQtGraph

PyQtGraph has modules to allow visual programming with flowcharts (and is used in the, somewhat related, ACQ4 Neurophysiology Acquisition and Analysis System:

OpenViBE Designer

OpenViBE Designer offers a visual programming environment for creating EEG analysis pipelines:

Orange

Another relevant tool is Orange:

Open source machine learning and data visualization for novice and expert. Interactive data analysis workflows with a large toolbox.

after installing 0.5.1 on Ubuntu I got following error:

Traceback (most recent call last):
  File "/home/daniel/.local/bin/mnelab", line 8, in <module>
    sys.exit(main())
  File "/home/daniel/.local/lib/python3.6/site-packages/mnelab/__main__.py", line 27, in main
    mp.set_start_method("spawn")  # required for Linux/macOS
  File "/usr/lib/python3.6/multiprocessing/context.py", line 242, in set_start_method
    raise RuntimeError('context has already been set')
RuntimeError: context has already been set

There is no history for actions done when plotting data (such as setting bads channels) either when using the "channel properties" dialog.

That way people can see which packages are optionally supported and which of them are currently installed.

To make the model module shorter, we could factor out all IO-related stuff (i.e. _load_XXX functions) including the special treatment of XDF.

When I supply invalid epoching values (e.g. tmin > tmax) in the dialog, an error dialog is shown. However, a bunch of strange unreadable text is also printed in the main window:

I have no idea why this is happening and I'd appreciate any help.

In some situations, using a GUI is just slower than typing a command, for example when renaming a bunch of channels according to a specific scheme. In such cases, the ability to enter Python commands would be very useful.

BrainVision recordings typically consist of three data files:

• the actual EEG data, .eeg
• the event marker file, .vmrk
• a header file, .vhdr, that contains references to the EEG and event marker files

For loading the EEG data into MNE, one passes the location of the .vhdr file to mne.io.read_raw_brainvision(). In case a dataset is incomplete (i.e., a .vmrk or .eeg file is missing / cannot be found), MNELAB crashes, as it doesn't handle the exception raised by MNE:

Extracting parameters from /Users/hoechenberger/Google Drive/TaLa/EEG_Analysis_outdated/1_DataEEG/Sub_721_071217/TaLa_721_01.vhdr...
Traceback (most recent call last):
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/mainwindow.py", line 322, in open_raw
    self.model.load(fname)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 38, in wrapper
    f(*args, **kwargs)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 119, in load
    raw = self._load_brainvision(fname)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 141, in _load_brainvision
    raw = mne.io.read_raw_brainvision(fname, preload=True)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 833, in read_raw_brainvision
    stim_channel=stim_channel, verbose=verbose)
  File "</Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/externals/decorator.py:decorator-gen-146>", line 2, in __init__
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/utils/_logging.py", line 89, in wrapper
    return function(*args, **kwargs)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 82, in __init__
    orig_units) = _get_vhdr_info(vhdr_fname, eog, misc, scale, montage)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 479, in _get_vhdr_info
    with open(mrk_fname, 'r') as tmp_mrk_f:
FileNotFoundError: [Errno 2] No such file or directory: '/Users/hoechenberger/Google Drive/TaLa/EEG_Analysis_outdated/1_DataEEG/Sub_721_071217/TaLa_721_01.vmrk'
[1]    3283 abort      mnelab

As mentionned in #34 , I could work on the handling of Epochs and Evoked data.

A suggestion for the changes in the structure of the code would be

• Change model.current['raw'] to model.current['data']
• Add a new keyword model.current['type'] set to 'raw', 'epochs' or 'evoked' depending on the type of data loaded.
• Add an epoching window close to

@cbrnr Clemens! :)

Do you have a few sample .bdf / .edf files for me to play around with?

The plot PSD dialog could use some options, e.g. whether or not to exclude bad channels.

The icons added in #61 cannot be found by MNELAB on startup, apparently the specified search path is not correct:

$ mnelab
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-folder_open-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-folder_open-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-view_list-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-view_list-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-waves-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-waves-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-poll-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-poll-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-blur_on-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-blur_on-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-filter_list-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-filter_list-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-youtube_searched_for-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-youtube_searched_for-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-grain-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-grain-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-view_stream-24px.svg', because: No such file or directory
qt.svg: Cannot open file '/Users/hoechenberger/Development/mnelab/images/outline-view_stream-24px.svg', because: No such file or directory

This is with master, I installed via pip install . and ran MNELAB via python -m mnelab

EDIT
Changed the working directory, and apparently MNELAB is looking for an the icons in the current working directory (to be specific, in an images/ subdir in the current working directory). Platform is macOS.

As mentioned in #34, a dialog window showing the command history (with an option to copy all or individual lines) would be great.

Currently, .fif.gz files cannot be loaded because (1) the file filter prevents users from selecting such files, and (2) dragging and dropping such files onto MNELAB crashes the app.

I cloned the mnelab directly from github. I installed mne and other packages through Anaconda under 'mne' environment. When I run mnelab.py as intructed, I faced the following error. Any quick fix?

(mne) F:\GitHub-Projects\mnelab\mnelab>python mnelab.py
Traceback (most recent call last):
File "mnelab.py", line 16, in
from .dialogs.filterdialog import FilterDialog
ModuleNotFoundError: No module named 'main.dialogs'; 'main' is not a package

In the data recorded using our BrainProducts system, there is no separate "stim channel", and I extract the events via

mne.find_events(raw_data)

This is not possible using MNELAB, as the "Find events..." dialog insists I select a "stim channel", which doesn't make sense in my case. There should be the option to leave the stim channel blank, in which case mne.find_events() would be called with stim_channel=None.

I'm sorry, the original issue description doesn't make sense anymore with recent versions of mne-python. Here's an updated, corrected description:

The events of BrainVision data have to be generated from the annotations of the raw data:

import mne

data = mne.io.read_raw_brainvision('eeg_recording.vhdr')
events = mne.events_from_annotations(data)

This work flow is currently not supported in MNELAB.

when updating to 5.0 on Ubuntu 18.04 and python3.6

   ERROR: Command errored out with exit status 1:
     command: /usr/bin/python3 -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/setup.py'"'"'; __file__='"'"'/tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/pip-egg-info
         cwd: /tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/
    Complete output (15 lines):
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "/tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/setup.py", line 25, in <module>
        extra_link_args=["-framework", "CoreFoundation"],
      File "/tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/pyobjc_setup.py", line 408, in Extension
        os_level = get_os_level()
      File "/tmp/pip-install-543k_rpn/pyobjc-framework-Cocoa/pyobjc_setup.py", line 218, in get_os_level
        pl = plistlib.readPlist("/System/Library/CoreServices/SystemVersion.plist")
      File "/usr/lib/python3.6/plistlib.py", line 162, in readPlist
        with _maybe_open(pathOrFile, 'rb') as fp:
      File "/usr/lib/python3.6/contextlib.py", line 81, in __enter__
        return next(self.gen)
      File "/usr/lib/python3.6/plistlib.py", line 120, in _maybe_open
        with open(pathOrFile, mode) as fp:
    FileNotFoundError: [Errno 2] No such file or directory: '/System/Library/CoreServices/SystemVersion.plist'
    ----------------------------------------
ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.

``

Documentation is urgently needed. Ideally, this should be a dedicated website describing how to install, start, and interact with MNELAB.

When loading a BrainVision file (*.vhdr), the size on disk shown is the size of the .vhdr file. However, this is not correct, because the EEG data resides in a file with the same base name but extension .eeg. The info widget should correctly show the size of the .eeg file.

Currently, only the subitems in the Export menu are disabled, but if possible the entire menu should be grayed out instead (not sure if it is possible though).

Dialogs should have tooltips for all fields, i.e. when hovering over a field a short explanation should pop up.

An MNELAB user reported that the app crashes when creating some annotations and then exporting to BrainVision format.

Used Annotations descriptions: ['1', '2', '3', '4', '8']
Traceback (most recent call last):
  File "C:\Python\Anaconda\envs\mnelab\lib\site-packages\mnelab\mainwindow.py", line 431, in export_file
    f(fname, ffilter)
  File "C:\Python\Anaconda\envs\mnelab\lib\site-packages\mnelab\model.py", line 253, in export_data
    self._export_bv(fname)
  File "C:\Python\Anaconda\envs\mnelab\lib\site-packages\mnelab\model.py", line 340, in _export_bv
    events = np.column_stack([events[:, [0, 2]], dur.astype(int)])
  File "<__array_function__ internals>", line 6, in column_stack
  File "C:\Python\Anaconda\envs\mnelab\lib\site-packages\numpy\lib\shape_base.py", line 641, in column_stack
    return _nx.concatenate(arrays, 1)
  File "<__array_function__ internals>", line 6, in concatenate
ValueError: all the input array dimensions for the concatenation axis must match exactly, but along dimension 0, the array at index 0 has size 90 and the array at index 1 has size 93

I tried to open a file via Open -> Recent; however that file had recently been moved to a different directory, so didn't exist in its original location anymore. MNELAB just crashes in this case, where it should handle the FileNotFoundError exception:

QWidget::insertAction: Attempt to insert null action
Extracting parameters from /Users/hoechenberger/Google Drive/TaLa/EEG_Analysis_outdated/1_DataEEG/Sub_902_211117/TaLa_902_01.vhdr...
Traceback (most recent call last):
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/mainwindow.py", line 652, in _load_recent
    self.open_raw(fname=action.text())
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/mainwindow.py", line 322, in open_raw
    self.model.load(fname)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 38, in wrapper
    f(*args, **kwargs)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 119, in load
    raw = self._load_brainvision(fname)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mnelab/model.py", line 141, in _load_brainvision
    raw = mne.io.read_raw_brainvision(fname, preload=True)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 833, in read_raw_brainvision
    stim_channel=stim_channel, verbose=verbose)
  File "</Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/externals/decorator.py:decorator-gen-146>", line 2, in __init__
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/utils/_logging.py", line 89, in wrapper
    return function(*args, **kwargs)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 82, in __init__
    orig_units) = _get_vhdr_info(vhdr_fname, eog, misc, scale, montage)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 449, in _get_vhdr_info
    settings, cfg, cinfostr, info = _aux_vhdr_info(vhdr_fname)
  File "/Users/hoechenberger/miniconda3/envs/mnelab/lib/python3.7/site-packages/mne/io/brainvision/brainvision.py", line 348, in _aux_vhdr_info
    with open(vhdr_fname, 'rb') as f:
FileNotFoundError: [Errno 2] No such file or directory: '/Users/hoechenberger/Google Drive/TaLa/EEG_Analysis_outdated/1_DataEEG/Sub_902_211117/TaLa_902_01.vhdr'
[1]    3235 abort      mnelab

It is possible to run ICA without a montage. However, when loading a montage afterwards, it should be possible to plot IC topoplots, but currently MNELAB crashes because the montage is not propagated to the ICA object. Maybe something to fix in MNE directly.

Currently, the history is a list of all commands that were executed in the GUI. However, when switching back and forth between different datasets, it might make more sense to have a per-dataset history.

The dialog asking to overwrite the current dataset should be more explicit, i.e. it should mention that the alternative to overwriting is creating a new dataset.

As mentioned in #34, and way to use ica.apply() would be a nice feature.

This could be a dialog showing all currently opened raw objects, so users first need to open all desired files, which they can then concatenate in the dialog window. This creates a new dataset.

MNELAB should support both PyQt5 and PySide2, either via a simple manual abstraction layer or via QtPy.

If a file that has been opened with MNELAB (i.e. it shows up in the list on the left side) is deleted from disk, MNELAB crashes. This also happens if e.g. I open files A, B, and C, then activate file A in the list, delete file B from disk, and then switch to file B in the list.

Currently MNELAB can convert annotations to events, but not the other way round.

Editing events and/or annotations should have history entries.

I just wanted to point out two related projects that mnelab might benefit from:

• eeg-timeFreqToolbox - seems to be a GUI for multiple time-frequency options that are in mne
• eeg-ICAtoolbox - judging from the name - some GUI tools for ICA.

I didn't investigate any of these projects beyond what is shown on the main github page, but thought it might be useful to notify you.

Cropping a raw object is pretty essential (and should be straightforward to implement because we only need to know tmin and tmax).

At some point we should add tests - pytest-qt should work nicely (I hope).

Maybe it would be nice to have a toolbar (with the option to show/hide it).

Currently it is not possible to see which export format MNELAB supports. To fix this, I suggest to create a submenu as follows:

Export data -> Export to FIF...
               Export to EEGLAB SET...
               Export to EDF/BDF...

This will be different for XDF and other MNE-supported file formats.

Using QFormLayout could simplify many dialogs, so MNELAB should use that when appropriate (instead of nested QHBoxLayout/QVBoxLayout/QGridLayout).

Hello all,
Can MNELAB open .gdf files? Are there certain files that it cannot open? I tries to open a dataset and it close the program instead of giving me an error that said it cannot open the file. Any help on this is appreciated.
Thanks.

Currently, users can close an open data set even when the plot window is open. Upon closing the plot window, MNELAB crashes because the data set has been closed.

Now that an initial release is available, it's time to track changes.

It seems the name mnelab has recently been adopted by another project, too: https://github.com/mohirio/mnelab

@cbrnr‘s mnelab went public first, though…

We should add a detailed guide for contributors (code style, how to make PRs, etc).