Numpy: Transform sparse matrix to ndarray

Question

I really couldn't google it. How to transform sparse matrix to ndarray?

Assume, I have sparse matrix t of zeros. Then

g = t.todense()
g[:10] 

matrix([[0],
    [0],
    [0],
    [0],
    [0],
    [0],
    [0],
    [0],
    [0],
    [0]])

instead of [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

Solution:

t.toarray().flatten()

ford · Accepted Answer · 2014-07-15 12:54:10Z

3

>>> a = np.asarray(g)
>>> a
array([[0],
       [0],
       [0],
       [0],
       [0],
       [0],
       [0],
       [0],
       [0],
       [0]])

Where g is your dense matrix in the example (after calling t.todense()).

You specifically asked for the output of

[0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

which has only one dimension. To get that, you'll want to flatten the array:

>>> flat_array = np.asarray(g).flatten()
>>> flat_array
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0])

Edit:

You can skip straight to the array from the sparse matrix with:

a = t.toarray()

answered Jul 15, 2014 at 11:23

ford

12k4 gold badges50 silver badges58 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

ZJS · Accepted Answer · 2014-07-15 11:30:36Z

0

Transpose your matrix to convert first column to first row

g = g.T

answered Jul 15, 2014 at 11:30

ZJS

4,0812 gold badges18 silver badges23 bronze badges