def lsa_solve_scipy(costs):
"""Solves the LSA problem using the scipy library."""
from scipy.optimize import linear_sum_assignment as scipy_solve
# scipy (1.3.3) does not support nan or inf values
finite_costs = add_expensive_edges(costs)
rids, cids = scipy_solve(finite_costs)
rids, cids = _exclude_missing_edges(costs, rids, cids)
return rids, cids
def lsa_solve_scipy(costs):
"""Solves the LSA problem using the scipy library."""
from scipy.optimize import linear_sum_assignment as scipy_solve
# Note there is an issue in scipy.optimize.linear_sum_assignment where
# it runs forever if an entire row/column is infinite or nan. We therefore
# make a copy of the distance matrix and compute a safe value that indicates
# 'cannot assign'. Also note + 1 is necessary in below inv-dist computation
# to make invdist bigger than max dist in case max dist is zero.
inv = ~np.isfinite(costs)
if inv.any():
costs = costs.copy()
valid = costs[~inv]
INVDIST = 2 * valid.max() + 1 if valid.shape[0] > 0 else 1.
costs[inv] = INVDIST
return scipy_solve(costs)