scipy.sparse.linalg.

spbandwidth#

scipy.sparse.linalg.spbandwidth(A)[source]#

Return the lower and upper bandwidth of a 2D numeric array.

Computes the lower and upper limits on the bandwidth of the sparse 2D array A. The result is summarized as a 2-tuple of positive integers (lo, hi). A zero denotes no sub/super diagonal entries on that side (triangular). The maximum value for lo (hi) is one less than the number of rows(cols).

Only the sparse structure is used here. Values are not checked for zeros.

Parameters:

ASciPy sparse array or matrix: A sparse matrix preferrably in CSR or CSC format.

Returns:

below, above2-tuple of int: The distance to the farthest non-zero diagonal below/above the main diagonal.

Added in version 1.15.0.

Examples

>>> import numpy as np
>>> from scipy.sparse.linalg import spbandwidth
>>> from scipy.sparse import csc_array, eye_array
>>> A = csc_array([[3, 0, 0], [1, -1, 0], [2, 0, 1]], dtype=float)
>>> spbandwidth(A)
(2, 0)
>>> D = eye_array(3, format='csr')
>>> spbandwidth(D)
(0, 0)