Added function `preprocess_data()` to be imported from `ecog.signal_processing`

[luiztauffer]

• Preprocess (resamples, filters) the raw data provided in a NWB file
• Runs spectral decomposition for bands in utils/bands.py
• Stores the processed signals and spectral decomposition power values in the same NWB file

[JesseLivezey]

@luiztauffer @bendichter FYI, we're not planning on maintaining anything in this repo. It's likely that in a few weeks we'll make the current version of our preprocessing code public, which was initially based on the code here. Were you looking to build substantially on this code?

[luiztauffer]

Hi @JesseLivezey thanks for the heads up! We’re altering the current process_ecog for two main reasons: 1 – make the preprocessing function directly usable by an ECoG visualization project. 2 – make it store the results in NWB files

Would you mind sharing what will be the main changes with the new software? =)

[JesseLivezey]

It might be easier to just copy the preprocessing functions and modify them in your repo. I could also make a different branch and merge them here (I'm using the docker branch for something specific).

The low-level changes to the new version are mostly for operating on arrays that are too large to process in memory all at once. Otherwise, they are the same. The high-level changes are to make the later stages of processing NWB compatible (and also somewhat specific to our lab).

[JesseLivezey]

I can't commit to managing anything in this repo, so it might be better for the other project not to depend on anything here.