Function test_partial_distance_duration_matrix

tests/test_Model.py:408–435

Tests that adding a full distance or duration matrix is not required. Any "missing" arcs are given default values.

(missing_value)

@pytest.mark.parametrize("missing_value", [5, 100, MAX_VALUE, MAX_VALUE + 1])
def test_partial_distance_duration_matrix(missing_value):
    """
    Tests that adding a full distance or duration matrix is not required. Any
    "missing" arcs are given default values.
    """
    model = Model()
    model.add_vehicle_type()

    depot = model.add_depot(0, 0)
    clients = [model.add_client(0, 1), model.add_client(1, 0)]

    # Only a subset of all possible edges exist. The model should be able to
    # handle that.
    model.add_edge(depot, clients[0], distance=1)
    model.add_edge(clients[0], clients[1], distance=2)
    model.add_edge(clients[1], depot, distance=1)

    # Edges that were not explicitly set should default to the missing value
    # argument, or MAX_VALUE, whichever is smaller.
    data = model.data(missing_value)
    distances = data.distance_matrix(profile=0)
    assert_equal(distances[0, 2], min(missing_value, MAX_VALUE))
    assert_equal(distances[1, 0], min(missing_value, MAX_VALUE))

    res = model.solve(MaxIterations(100), seed=4, missing_value=missing_value)
    assert_equal(res.best.num_routes(), 1)
    assert_equal(res.cost(), 4)  # depot -> client 1 -> client 2 -> depot
    assert_(res.is_feasible())


def test_data_warns_about_scaling_issues(recwarn):