find stable matching of minimal distance between two sets of points

stable_matching_between_2_sets_of_points.py

import numpy as np
import scipy.optimize

a = np.array([[0, 0], [1, 1], [2, 2]])
b = np.array([[0, 1], [2, 2.2], [1, 1.2], [0, 0]])

indices = np.indices((len(a), len(b)))
# array([[[0, 0, 0, 0],
#         [1, 1, 1, 1],
#         [2, 2, 2, 2]],
#
#        [[0, 1, 2, 3],
#         [0, 1, 2, 3],
#         [0, 1, 2, 3]]])

distance_matrix = np.vectorize(lambda i, j: np.linalg.norm(a[i] - b[j]))\
        (*indices)
# array([[1.  , 2.97, 1.56, 0.  ],
#        [1.  , 1.56, 0.2 , 1.41],
#        [2.24, 0.2 , 1.28, 2.83]])

ii, jj = scipy.optimize.linear_sum_assignment(distance_matrix)

# (array([0, 1, 2]), array([3, 2, 1]))