pyriemann is a python package for covariance matrices manipulation and classification through riemannian geometry.

The primary target is classification of multivariate biosignals, like EEG, MEG or EMG.

This is work in progress ... stay tuned.

This code is BSD-licenced (3 clause).

The documentation is available on http://pythonhosted.org/pyriemann

pip install pyriemann

For the latest version, you can install the package from the sources using the setup.py script

python setup.py install

Most of the functions mimic the scikit-learn API, and therefore can be directly used with sklearn. For example, for cross-validation classification of EEG signal using the MDM algorithm described in [4] , it is easy as :

import pyriemann
from sklearn.cross_validation import cross_val_score

# load your data
X = ... # your EEG data, in format Ntrials x Nchannels X Nsamples
y = ... # the labels

# estimate covariances matrices
cov = pyriemann.estimation.Covariances().fit_transform(X)

# cross validation
mdm = pyriemann.classification.MDM()

accuracy = cross_val_score(mdm,cov,y)

print(accuracy.mean())

You can also pipeline methods using sklearn Pipeline framework. For example, to classify EEG signal using a SVM classifier in the tangent space, described in [5] :

from pyriemann.estimation import Covariances
from pyriemann.tangentspace import TangentSpace

from sklearn.pipeline import make_pipeline
from sklearn.svm import SVC
from sklearn.cross_validation import cross_val_score

# load your data
X = ... # your EEG data, in format Ntrials x Nchannels X Nsamples
y = ... # the labels

# build your pipeline
covest = Covariances()
ts = TangentSpace()
svc = SVC(kernel='linear')

clf = make_pipeline(covest,ts,svc)
# cross validation
accuracy = cross_val_score(clf,X,y)

print(accuracy.mean())

Check out the example folder for more examples !

If you make a modification, run the test suite before submitting a pull request

nosetests

The package aims at adopting the Scikit-Learn and MNE-Python conventions as much as possible. See their contribution guidelines before contributing to the repository.

[1] A. Barachant, M. Congedo ,"A Plug&Play P300 BCI Using Information Geometry", arXiv:1409.0107. link

[2] M. Congedo, A. Barachant, A. Andreev ,"A New generation of Brain-Computer Interface Based on Riemannian Geometry", arXiv: 1310.8115. link

[3] A. Barachant and S. Bonnet, "Channel selection procedure using riemannian distance for BCI applications," in 2011 5th International IEEE/EMBS Conference on Neural Engineering (NER), 2011, 348-351. pdf

[4] A. Barachant, S. Bonnet, M. Congedo and C. Jutten, “Multiclass Brain-Computer Interface Classification by Riemannian Geometry,” in IEEE Transactions on Biomedical Engineering, vol. 59, no. 4, p. 920-928, 2012. pdf

[5] A. Barachant, S. Bonnet, M. Congedo and C. Jutten, “Classification of covariance matrices using a Riemannian-based kernel for BCI applications“, in NeuroComputing, vol. 112, p. 172-178, 2013. pdf

• Improved documentation
• Added TSclassifier for out-of the box tangent space classification.
• Added Wasserstein distance and mean.
• Added NearestNeighbor classifier.
• Added Softmax probabilities for MDM.
• Added CSP for covariance matrices.
• Added Approximate Joint diagonalization algorithms (JADE, PHAM, UWEDGE).
• Added ALE mean.
• Added Multiclass CSP.
• API: param name changes in CospCovariances to comply to Scikit-Learn.
• API: attributes name changes in most modules to comply to the Scikit-Learn naming convention.
• Added HankelCovariances estimation
• Added SPoC spatial filtering
• Added Harmonic mean
• Added Kullback leibler mean

• Added multiprocessing for MDM with joblib
• Added kullback-leibler divergence
• Added Riemannian Potato
• Added sample_weight for mean estimation and MDM