Coverage for /usr/lib/python3/dist-packages/scipy/sparse/_matrix_io.py : 26%

from __future__ import division, print_function, absolute_import 

import sys 

import numpy as np 

import scipy.sparse 

from scipy._lib._version import NumpyVersion 

__all__ = ['save_npz', 'load_npz'] 

if NumpyVersion(np.__version__) >= '1.10.0': 

# Make loading safe vs. malicious input 

PICKLE_KWARGS = dict(allow_pickle=False) 

else: 

PICKLE_KWARGS = dict() 

def save_npz(file, matrix, compressed=True): 

""" Save a sparse matrix to a file using ``.npz`` format. 

Parameters 

---------- 

file : str or file-like object 

Either the file name (string) or an open file (file-like object) 

where the data will be saved. If file is a string, the ``.npz`` 

extension will be appended to the file name if it is not already 

there. 

matrix: spmatrix (format: ``csc``, ``csr``, ``bsr``, ``dia`` or coo``) 

The sparse matrix to save. 

compressed : bool, optional 

Allow compressing the file. Default: True 

See Also 

-------- 

scipy.sparse.load_npz: Load a sparse matrix from a file using ``.npz`` format. 

numpy.savez: Save several arrays into a ``.npz`` archive. 

numpy.savez_compressed : Save several arrays into a compressed ``.npz`` archive. 

Examples 

-------- 

Store sparse matrix to disk, and load it again: 

>>> import scipy.sparse 

>>> sparse_matrix = scipy.sparse.csc_matrix(np.array([[0, 0, 3], [4, 0, 0]])) 

>>> sparse_matrix 

<2x3 sparse matrix of type '<class 'numpy.int64'>' 

with 2 stored elements in Compressed Sparse Column format> 

>>> sparse_matrix.todense() 

matrix([[0, 0, 3], 

[4, 0, 0]], dtype=int64) 

>>> scipy.sparse.save_npz('/tmp/sparse_matrix.npz', sparse_matrix) 

>>> sparse_matrix = scipy.sparse.load_npz('/tmp/sparse_matrix.npz') 

>>> sparse_matrix 

<2x3 sparse matrix of type '<class 'numpy.int64'>' 

with 2 stored elements in Compressed Sparse Column format> 

>>> sparse_matrix.todense() 

matrix([[0, 0, 3], 

[4, 0, 0]], dtype=int64) 

""" 

arrays_dict = {} 

if matrix.format in ('csc', 'csr', 'bsr'): 

arrays_dict.update(indices=matrix.indices, indptr=matrix.indptr) 

elif matrix.format == 'dia': 

arrays_dict.update(offsets=matrix.offsets) 

elif matrix.format == 'coo': 

arrays_dict.update(row=matrix.row, col=matrix.col) 

else: 

raise NotImplementedError('Save is not implemented for sparse matrix of format {}.'.format(matrix.format)) 

arrays_dict.update( 

format=matrix.format.encode('ascii'), 

shape=matrix.shape, 

data=matrix.data 

) 

if compressed: 

np.savez_compressed(file, **arrays_dict) 

else: 

np.savez(file, **arrays_dict) 

def load_npz(file): 

""" Load a sparse matrix from a file using ``.npz`` format. 

Parameters 

---------- 

file : str or file-like object 

Either the file name (string) or an open file (file-like object) 

where the data will be loaded. 

Returns 

------- 

result : csc_matrix, csr_matrix, bsr_matrix, dia_matrix or coo_matrix 

A sparse matrix containing the loaded data. 

Raises 

------ 

IOError 

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

See Also 

-------- 

scipy.sparse.save_npz: Save a sparse matrix to a file using ``.npz`` format. 

numpy.load: Load several arrays from a ``.npz`` archive. 

Examples 

-------- 

Store sparse matrix to disk, and load it again: 

>>> import scipy.sparse 

>>> sparse_matrix = scipy.sparse.csc_matrix(np.array([[0, 0, 3], [4, 0, 0]])) 

>>> sparse_matrix 

<2x3 sparse matrix of type '<class 'numpy.int64'>' 

with 2 stored elements in Compressed Sparse Column format> 

>>> sparse_matrix.todense() 

matrix([[0, 0, 3], 

[4, 0, 0]], dtype=int64) 

>>> scipy.sparse.save_npz('/tmp/sparse_matrix.npz', sparse_matrix) 

>>> sparse_matrix = scipy.sparse.load_npz('/tmp/sparse_matrix.npz') 

>>> sparse_matrix 

<2x3 sparse matrix of type '<class 'numpy.int64'>' 

with 2 stored elements in Compressed Sparse Column format> 

>>> sparse_matrix.todense() 

matrix([[0, 0, 3], 

[4, 0, 0]], dtype=int64) 

""" 

with np.load(file, **PICKLE_KWARGS) as loaded: 

try: 

matrix_format = loaded['format'] 

except KeyError: 

raise ValueError('The file {} does not contain a sparse matrix.'.format(file)) 

matrix_format = matrix_format.item() 

if sys.version_info[0] >= 3 and not isinstance(matrix_format, str): 

# Play safe with Python 2 vs 3 backward compatibility; 

# files saved with Scipy < 1.0.0 may contain unicode or bytes. 

matrix_format = matrix_format.decode('ascii') 

try: 

cls = getattr(scipy.sparse, '{}_matrix'.format(matrix_format)) 

except AttributeError: 

raise ValueError('Unknown matrix format "{}"'.format(matrix_format)) 

if matrix_format in ('csc', 'csr', 'bsr'): 

return cls((loaded['data'], loaded['indices'], loaded['indptr']), shape=loaded['shape']) 

elif matrix_format == 'dia': 

return cls((loaded['data'], loaded['offsets']), shape=loaded['shape']) 

elif matrix_format == 'coo': 

return cls((loaded['data'], (loaded['row'], loaded['col'])), shape=loaded['shape']) 

else: 

raise NotImplementedError('Load is not implemented for ' 

'sparse matrix of format {}.'.format(matrix_format))