Principal component analysis (source code)

= Principal component analysis
{wiki}

Principal component analysis finds orthogonal directions of greatest sample variance. Its first loading vector is a unit eigenvector corresponding to the largest eigenvalue of the sample covariance matrix.