Principal component analysis in Python

Question

I\'d like to use principal component analysis (PCA) for dimensionality reduction. Does numpy or scipy already have it, or do I have to roll my own using numpy.linalg.eigh?<

Answer 1

You can quite easily "roll" your own using scipy.linalg (assuming a pre-centered dataset data):

covmat = data.dot(data.T)
evs, evmat = scipy.linalg.eig(covmat)

Then evs are your eigenvalues, and evmat is your projection matrix.

If you want to keep d dimensions, use the first d eigenvalues and first d eigenvectors.

Given that scipy.linalg has the decomposition and numpy the matrix multiplications, what else do you need?