Singular value decomposition (SVD), low-rank approximation (LRA) in Python

If you want to speed up, I don't think you will use it from Lapack or Python, so I hope you can use such scientific and technological calculations as much as the answers of people who do it in cuda, C, or C ++.

I thought it was a memo

Source code

low_rank_approximation.py

import numpy as np
from scipy import linalg

def low_rank_approximation(a,rank):
    u, s, v = linalg.svd(a)
    ur = u[:, :rank]
    sr = np.matrix(linalg.diagsvd(s[:rank], rank,rank))
    vr = v[:rank, :]
    return np.asarray(ur*sr*vr)

A = np.array([[1,2,3],[4,5,6],[7,8,9]])
print A
B = low_rank_approximation(A,1)
print B

# python low_rank_approximation.py
# [[1 2 3]
#  [4 5 6]
#  [7 8 9]]
# [[ 1.73621779  2.07174246  2.40726714]
#  [ 4.2071528   5.02018649  5.83322018]
#  [ 6.6780878   7.96863051  9.25917322]]

reference

• http://qiita.com/kaityo256/items/48de63526b469235d16a
• http://qiita.com/kyoro1/items/4df11e933e737703d549
• http://oppython.hatenablog.com/entry/2014/01/21/003245