I am trying to get the condition number of a scipy sparse matrix. The way I managed to do this is to convert the matrix to a dense one, and then get its eigenvalues:
$ python
Python 3.5.2 (v3.5.2:4def2a2901a5, Jun 26 2016, 10:47:25)
[GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from numpy import array
>>> import numpy as np
>>> import scipy.sparse as sparse
>>> I = array([0,3,1,0])
>>> J = array([0,3,1,2])
>>> V = array([4,5,7,9])
>>> A = sparse.coo_matrix((V,(I,J)),shape=(4,4))
>>> A = A.todense()
>>> eig = np.linalg.eig(A)
>>> eig = eig[0].real, np.array(eig[1].real)
>>> def split(array, cond):
... return (array[cond], array[~cond])
...
>>> eigv, zero = split(eig[0], eig[0]>1e-10)
>>> cond = max(eigv) / min(eigv)
>>> cond
1.75
As expected, this becomes impracticable for large matrices. I was wondering how is this done correctly in Python?
source
share