`scipy.spatial._kdtree:KDTree.sparse_distance_matrix`

source: /scipy/spatial/_kdtree.py:886

Signature

def sparse_distance_matrix(self, other, max_distance, p = 2.0, output_type = dok_matrix)

Summary

Compute a sparse distance matrix.

Extended Summary

Computes a distance matrix between two KDTrees, leaving as zero any distance greater than max_distance.

Parameters

other : KDTree: The other KDTree to compute distances against.
max_distance : positive float: Maximum distance within which neighbors are returned. Distances above this value are returned as zero.
p : float, 1<=p<=infinity: Which Minkowski p-norm to use. A finite large p may cause a ValueError if overflow can occur.
output_type : str, optional: Which container to use for output data. Options: 'dok_array', 'coo_array', 'dict', or 'ndarray'. Legacy options 'dok_matrix' and 'coo_matrix' are still available. Default: 'dok_matrix'.
warning
dok_matrix and coo_matrix are being replaced. All new code using scipy sparse should use sparse array types 'dok_array' or 'coo_array'. The default value of output_type will be deprecated at v1.19 and switch from 'dok_matrix' to 'dok_array' in v1.21. The values 'dok_matrix' and 'coo_matrix' continue to work, but will go away eventually.
versionadded 1.6.0

Returns

result : dok_array, coo_array, dict or ndarray: Sparse matrix representing the results in "dictionary of keys" format. If a dict is returned the keys are (i,j) tuples of indices. If output_type is 'ndarray' a record array with fields 'i', 'j', and 'v' is returned,

Examples

You can compute a sparse distance matrix between two kd-trees:

import numpy as np
from scipy.spatial import KDTree
rng = np.random.default_rng()
points1 = rng.random((5, 2))
points2 = rng.random((5, 2))
kdtree1 = KDTree(points1)
kdtree2 = KDTree(points2)
sdm = kdtree1.sparse_distance_matrix(kdtree2, 0.3, output_type="dok_array")

✓

sdm.toarray()

✗

array([[0.     , 0.     , 0.27296, 0.2546 , 0.     ],
       [0.06422, 0.     , 0.     , 0.     , 0.     ],
       [0.19659, 0.     , 0.     , 0.     , 0.     ],
       [0.13361, 0.     , 0.     , 0.     , 0.     ],
       [0.     , 0.     , 0.     , 0.     , 0.     ]])

You can check distances above the `max_distance` are zeros:

from scipy.spatial.distance import cdist

✓

cdist(points1, points2)

✗

array([[0.92051, 0.34091, 0.27296, 0.2546 , 0.38924],
       [0.06422, 0.59761, 0.93863, 0.7052 , 0.82604],
       [0.19659, 0.39908, 0.70936, 0.51702, 0.59573],
       [0.13361, 0.60411, 0.96603, 0.70487, 0.86201],
       [0.40768, 0.44538, 0.57588, 0.56199, 0.42595]])

Aliases

scipy.spatial.KDTree.sparse_distance_matrix