Package for modelling s/M/EEG responses to speech. In other words, for mapping speech features, through python (sPyeech) to EEG (sPyEEG)... and the other way around!
Not mind-reading for espionage purposes ;).
Package builds on top on MNE and relies on a similar set of dependencies and 3rd party packages listed in environment.yml
. You can easily set up the environment via Conda package manager by running in terminal From terminal (or conda shell on Windows):
conda env update --file environment.yml
Then activate the created environment by running:
conda activate spyeeg
To get the package installed only through symbolic links, namely so that you can modify the source code and use modified versions at will when importing the package in your python scripts do:
python setup.py develop
Otherwise, for a standard installation (but this will require to be installed if you need to install another version of the library):
python setup.py install
- macOS Big Sur v11.1
- Ubuntu 18.04.5 LTS
- Windows 10 22H2
- models - for all your modelling needs
- TRF: Temporal Response Function a.k.a Ridge regression a.k.a. fancy linear regression, optimized for speed
- iRRR: integrative reduced rank regression a.k.a fancier linear regression
- _methods: useful methods used by several model classes
- CCA/Decoder/ERP: to be properly written at a later date.
- feat - simple feature extraction.
- preproc - useful preprocessing functions (filters, detrending...etc)
- viz - visualization tools. To come.
- utils - misc.
Note: Sample data required for demos can be downloaded here. When downloaded place the files in the demos/Data
folder.
- Modelling (
models
)- Basic TRF modelling:
demo/Demo_TRF.py
- iRRR demo: coming soon
- Basic TRF modelling:
- Feature extraction (
feat
)- Speech envelope extraction:
demo/Demo_envelopes.py
- Speech envelope extraction:
- Pierre Guilleminot ([email protected])
- Mikolaj Kegler ([email protected])
Last updated: 15th Apr 2024