pyntbci.transformers.TRCA

class pyntbci.transformers.TRCA(n_components: int = 1)

Task related component analysis (TRCA). Maximizes the intra-class covariances, i.e., the intra-class consistency [2]. TRCA was applied to (SSVEP) BCI [3]. Alternative implementations, also used as example for this code, see Matlab code in [2] for the original, Matlab code in [4] for the SSVEP BCI introduction, and two Python implementation in MOABB [5], and MEEGKit [6].

Parameters:

n_components (int (default: 1)) – The number of TRCA components to use.

w_

The weight vector to project X of shape (n_features, n_components).

Type:

NDArray

References

[2] (1,2)

Tanaka, H., Katura, T., & Sato, H. (2013). Task-related component analysis for functional neuroimaging and application to near-infrared spectroscopy data. NeuroImage, 64, 308-327. doi: 10.1016/j.neuroimage.2012.08.044

[3]

Nakanishi, M., Wang, Y., Chen, X., Wang, Y. T., Gao, X., & Jung, T. P. (2017). Enhancing detection of SSVEPs for a high-speed brain speller using task-related component analysis. IEEE Transactions on Biomedical Engineering, 65(1), 104-112. doi: 10.1109/TBME.2017.2694818

[4]

https://github.com/mnakanishi/TRCA-SSVEP/blob/master/src/train_trca.m

[5]

https://github.com/NeuroTechX/moabb/blob/develop/moabb/pipelines/classification.py

[6]

https://github.com/nbara/python-meegkit/blob/master/meegkit/trca.py

fit(X: ndarray[tuple[int, ...], dtype[_ScalarType_co]], y: ndarray[tuple[int, ...], dtype[_ScalarType_co]] = None) → TransformerMixin

Fit TRCA.

Parameters:

• X (NDArray) – Data matrix of shape (n_trials, n_features, n_samples).

• y (NDArray (default: None)) – Not used.

Returns:

self – Returns the instance itself.

Return type:

TransformerMixin

transform(X: ndarray[tuple[int, ...], dtype[_ScalarType_co]], y: ndarray[tuple[int, ...], dtype[_ScalarType_co]] = None) → ndarray[tuple[int, ...], dtype[_ScalarType_co]]

Transform the data matrix from feature space to component space. Note, can operate on both X and y or just X. If X and y are provided, X is filtered with class-specific filters. If only X is provided and a multi-class filter was learned, all trials are filtered with all filters. If only one filter was learned, then only this filter is applied.

Parameters:

• X (NDArray) – Data matrix of shape (n_trials, n_features, n_samples).

• y (NDArray (default: None)) – Not used.

Returns:

X – Projected data matrix of shape (n_trials, n_components, n_samples).

Return type:

NDArray