scipy.sparse.

load_npz#

scipy.sparse.load_npz(file)[source]#

Load a sparse array/matrix from a file using .npz format.

Parameters:

filestr or file-like object: Either the file name (string) or an open file (file-like object) where the data will be loaded.

Returns:

resultcsc_array, csr_array, bsr_array, dia_array or coo_array: A sparse array/matrix containing the loaded data.

Raises:

OSError: If the input file does not exist or cannot be read.

See also

scipy.sparse.save_npz: Save a sparse array/matrix to a file using .npz format.
numpy.load: Load several arrays from a .npz archive.

Examples

Store sparse array/matrix to disk, and load it again:

>>> import numpy as np
>>> import scipy as sp
>>> sparse_array = sp.sparse.csc_array([[0, 0, 3], [4, 0, 0]])
>>> sparse_array
<Compressed Sparse Column sparse array of dtype 'int64'
    with 2 stored elements and shape (2, 3)>
>>> sparse_array.toarray()
array([[0, 0, 3],
       [4, 0, 0]], dtype=int64)

>>> sp.sparse.save_npz('/tmp/sparse_array.npz', sparse_array)
>>> sparse_array = sp.sparse.load_npz('/tmp/sparse_array.npz')

>>> sparse_array
<Compressed Sparse Column sparse array of dtype 'int64'
    with 2 stored elements and shape (2, 3)>
>>> sparse_array.toarray()
array([[0, 0, 3],
       [4, 0, 0]], dtype=int64)

In this example we force the result to be csr_array from csr_matrix >>> sparse_matrix = sp.sparse.csc_matrix([[0, 0, 3], [4, 0, 0]]) >>> sp.sparse.save_npz(‘/tmp/sparse_matrix.npz’, sparse_matrix) >>> tmp = sp.sparse.load_npz(‘/tmp/sparse_matrix.npz’) >>> sparse_array = sp.sparse.csr_array(tmp)