Using with SciPy¶

Import¶

In [1]:

import sparse

import numpy as np
import scipy.sparse as sps

import sparse import numpy as np import scipy.sparse as sps

Create Arrays¶

In [2]:

rng = np.random.default_rng(42)
M = 1_000
DENSITY = 0.01
a = sparse.random((M, M), density=DENSITY, format="csc")
identity = sparse.eye(M, format="csc")

rng = np.random.default_rng(42) M = 1_000 DENSITY = 0.01 a = sparse.random((M, M), density=DENSITY, format="csc") identity = sparse.eye(M, format="csc")

Invert and verify matrix¶

This showcases the scipy.sparse.linalg integration.

In [3]:

a_inv = sps.linalg.spsolve(a, identity)
np.testing.assert_array_almost_equal((a_inv @ a).todense(), identity.todense())

a_inv = sps.linalg.spsolve(a, identity) np.testing.assert_array_almost_equal((a_inv @ a).todense(), identity.todense())

Calculate the graph distances¶

This showcases the scipy.sparse.csgraph integration.

In [4]:

sps.csgraph.bellman_ford(sparse.eye(5, k=1) + sparse.eye(5, k=-1), return_predecessors=False)

sps.csgraph.bellman_ford(sparse.eye(5, k=1) + sparse.eye(5, k=-1), return_predecessors=False)

Out[4]:

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