`scipy.io._fast_matrix_market:mmread`

source: /scipy/io/_fast_matrix_market/__init__.py:301

Signature

def mmread(source, *, spmatrix = <object object at 0x0000>)

Summary

Reads the contents of a Matrix Market file-like 'source' into a matrix.

Parameters

source : str or file-like: Matrix Market filename (extensions .mtx, .mtz.gz) or open file-like object.
spmatrix : bool, optional (default: True): If True, return sparse matrix. Otherwise return sparse array.
deprecated 1.18.0
The default value for spmatrix is changing to False in v1.20. That means the default return value will be a sparse array. Unless you use * instead of @, ** for matrix power, or you depend on 2D shapes from e.g. A.sum(axis=0) it may not matter to you. See Migration from spmatrix to sparray.

Returns

a : ndarray or coo_array: Dense or sparse array depending on the matrix format in the Matrix Market file.

Notes

Examples

from io import StringIO
from scipy.io import mmread

✓

text = '''%%MatrixMarket matrix coordinate real general
 5 5 7
 2 3 1.0
 3 4 2.0
 3 5 3.0
 4 1 4.0
 4 2 5.0
 4 3 6.0
 4 4 7.0
'''

✓

``mmread(source)`` returns the data as sparse array in COO format.

m = mmread(StringIO(text), spmatrix=False)

✓

✗

<COOrdinate sparse array of dtype 'float64'
	with 7 stored elements and shape (5, 5)>

m.toarray()

✓

array([[0., 0., 0., 0., 0.],
       [0., 0., 1., 0., 0.],
       [0., 0., 0., 2., 3.],
       [4., 5., 6., 7., 0.],
       [0., 0., 0., 0., 0.]])

This method is threaded. The default number of threads is equal to the number of CPUs in the system. Use `threadpoolctl <https://github.com/joblib/threadpoolctl>`_ to override:

import threadpoolctl
with threadpoolctl.threadpool_limits(limits=2):
    m = mmread(StringIO(text), spmatrix=False)

⚠

Traceback (most recent call last):
  File "<doctest scipy.io._fast_matrix_market:mmread[0]>", line 1, in <module>
    with threadpoolctl.threadpool_limits(limits=2):
         ^^^^^^^^^^^^^
NameError: name 'threadpoolctl' is not defined

Aliases

scipy.io.mmread