Randomly shuffle a sparse matrix in python

Question

is there an easy way to shuffle a sparse matrix in python?

This is how I shuffle a non-sparse matrix:

    index = np.arange(np.shape(matrix)[0])
    np.random.shuffle(index)
    return matrix[index]

How can I do it with numpy sparse?

score 18 · Accepted Answer · answered Sep 01 '12 at 19:27

18

Ok, found it. The sparse format looks a bit confusing in the print-out.

    index = np.arange(np.shape(matrix)[0])
    print index
    np.random.shuffle(index)
    return matrix[index, :]

answered Sep 01 '12 at 19:27

Puckl

Works only for csr_matrix. I am getting an error 'coo_matrix' object has no attribute '__getitem__'. – Ramakrishnan Kannan Dec 16 '16 at 17:12
One should note that this creates a copy of the array, meaning that unless you have enough RAM to keep two copies in memory it won't work. – Thomas Ahle Feb 04 '20 at 08:43

score 1 · Answer 2 · edited May 23 '17 at 11:48

1

In case anyone is looking to randomly get a subsample of rows from a sparse matrix, this related post may also be useful: How should I go about subsampling from a scipy.sparse.csr.csr_matrix and a list

edited May 23 '17 at 11:48

Community

answered Mar 18 '16 at 14:11

shaneb

score 0 · Answer 3 · answered Jan 22 '19 at 22:46

0

A better way can be shuffling the index of CSR Matrix and fetching the rows of matrix as such:

from random import shuffle
indices = np.arange(matrix.shape[0]) #gets the number of rows 
shuffle(indices)
shuffled_matrix = matrix[list(indices)]

answered Jan 22 '19 at 22:46

Anurag Pandey

3 Answers3