Function to generate an SVD low-rank approximation of a matrix, using numpy.linalg.svd. Can be used as a form of compression, or to reduce the condition number of a matrix.

svd_approximate.py

import numpy as np


def low_rank_approx(SVD=None, A=None, r=1):
    """
    Computes an r-rank approximation of a matrix
    given the component u, s, and v of it's SVD

    Requires: numpy
    """
    if not SVD:
        SVD = np.linalg.svd(A, full_matrices=False)
    u, s, v = SVD
    Ar = np.zeros((len(u), len(v)))
    for i in xrange(r):
        Ar += s[i] * np.outer(u.T[i], v[i])
    return Ar

if __name__ == "__main__":
    """
    Test: visualize an r-rank approximation of `lena`
    for increasing values of r

    Requires: scipy, matplotlib
    """
    from scipy.misc import lena
    import pylab
    x = lena()
    u, s, v = np.linalg.svd(x, full_matrices=False)
    i = 1
    pylab.figure()
    pylab.ion()
    while i < len(x) - 1:
        y = low_rank_approx((u, s, v), r=i)
        pylab.imshow(y, cmap=pylab.cm.gray)
        pylab.draw()
        i += 1
        #print percentage of singular spectrum used in approximation
        print "%0.2f %%" % (100 * i / 512.)

I think Lena is not used anymore in scipy. I used "ascent." Posted link below. Thanks!

https://docs.scipy.org/doc/scipy/reference/generated/scipy.misc.ascent.html