[x] Create common_spatial_pattern.py: apply the Common Spatial algorithm on raw EEG/MEG trials
[x] Create data.py: recover raw EEG/MEG trials from .mat format, select wanted channels (EEG type and /or MEG type), create labels (number of spikes in the trial)
[x] Create dataloader.py: get random weighted dataloader for training (to correct class imbalance) and shuffle dataloader for testing
[x] Create utils.py: contains function to check balance and class distribution
[x] Create data_visualizaion.py: contains functions to visualize EEG/MEG trials (with or without CSP projection) thanks to https://mne.tools/stable/index.html
Description
This PR enables to apply the Common Spatial algorithm proposed in the model of
https://arxiv.org/pdf/2106.11170.pdf. It also enables to recover data from .mat format to numpy array with corresponding labels (number of spikes in EEG signals) and to produce dataloaders of data.
Checklist
Description
This PR enables to apply the Common Spatial algorithm proposed in the model of https://arxiv.org/pdf/2106.11170.pdf. It also enables to recover data from .mat format to numpy array with corresponding labels (number of spikes in EEG signals) and to produce dataloaders of data.