KamalakerDadi

Basically need some clarification on this: There is a file `test_eeglab_infomax.py` which needs to be ported. This file is specifically to test with scikit-learn results to compare with EEGLAB matlab...

After discussions with Denis, I am taking over this idea to port MNE-Python files to Scikit Learn. Links: Main codebase: https://github.com/mne-tools/mne-python/blob/master/mne/preprocessing/infomax_.py Tests: https://github.com/mne-tools/mne-python/blob/master/mne/preprocessing/tests/test_infomax.py Work goes like this: - I will...

@dengemann Do you think we need to change `l_rate` and to be prefixed to `l_rate=0.001` ? Current design: ``` python l_rate = 0.01 / math.log(n_features ** 2.0) ``` This has...

Ok. I will see it.

plot_prob_atlas displays all maps on a background image. does this reports also the timeseries extracted per map and subject? That would be interesting.

maps masking is done on 4D as input. list of 3D is still a 4D. thats why Jerome code works. But if you provide list of 4D, it should not...

Yes, PR will come soon. All masker families should benefit via joblib caching and parallelization and multi-subject extraction. This is part of issue opened at #1240.

Random Walker just returns the same if no markers are found. Probably, yes this might happen failing to find peak extraction. There are also many cases to consider as well...

> Maybe the special case should be done in the RegionExtractor, and not the random walker? No, not in random walker. > Also, in the logic of the function of...

>I also tried and failed to reproduce it with the development dataset: May be unrelated to what we are asking here. But, the t_r of development dataset is 2.0 #2021