PyNTBCI

The Python Noise-Tagging BCI toolbox (PyntBCI) contains custom software to setup a code-modulated evoked response BCI. For instance, it contains routines to create stimulus sequences (i.e., noise codes) and to analyse and decode code-modulated evoked potentials (e.g., c-VEP). The toolbox can be taken from GitHub here, or be installed directly from PyPI:

pip install pyntbci

Using PyntBCI is similar to using any estimator object from Scikit-Learn. For instance, the reconvolution CCA pipeline can be accessed by using the rCCA class and its fit and predict methods:

import pynt
rcca = pynt.classifiers.rCCA(codes=codes, fs=fs)
rcca.fit(X_trn, y_trn)
yh_tst = rcca.predict(X_tst)