Refactor the codebase so that I can generate automatically the raw figure for the paper about assessing-the-quality-of-wearable-eeg-systems-using-functional-connectivity

[yacineMahdid]

Good news I already have the spatially downsampled data for all the headset. This can be used directly to generate all the figure in one go.

[yacineMahdid]

TODO

• Modify NeuroAlgo Recording class to be able to load a .mat file as well as a .set file
• Setup the analysis pipeline to have the same parameters as we are using in the paper
• Run the full analysis remotely on the lab computer on all frequency bands

At this point I realised that I have to reorder all the headset, however I only made my reordering function work for the egi129. Will need to create a cassette for the other headset too otherwise this is going to be a mess. They are not ordered in a FTCPO order by default.

• Setup the figure generation code to churn through the /result folder
• Reproduce all the figures in the paper at the frequency band at alpha ( + full band for spectrogram)
• create the .csv with the FTCPO ordering for all headsets.
• test the reordering with all the headsets.
• make a module to only take the left or right part of the brain
• figure out a way to scale automatically by groups of four in the figures of all headsets or 2 in the egi full figure.

At this point I can automatically create the figures, I just need to automate it for wPLI and dPLI together without making the code a big mess and generating .png file that are exactly of the right size.

• figure out how to scale the figure to be of the approximate size we need to cut it and paste it in inkscape.
• figure out how to save the .png version of the figure instead of the .fig
• create the wPLI and dPLI figures at all frequency bands for all headset (including the downsampled version)
• figure out wtf should I do with 108 plots?

At this point the pipeline should be documented well enough so that I can apply it to generate the analysis for all the other frequency bands.