Eigen::MatrixXf covarianceMatrix(const Table<Eigen::VectorXf> &samples) {
size_t featuresDimension = samples.columnsNumber();
Eigen::MatrixXf samplesMatrix(samples.rowsNumber(), featuresDimension);
for (size_t i = 0; i < samples.rowsNumber(); ++i) {
samplesMatrix.row(i) = samples[i];
}
Eigen::MatrixXf centered = samplesMatrix.rowwise() - samplesMatrix.colwise().mean();
return (centered.adjoint() * centered) / double(samplesMatrix.rows());
}
