def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (1, 1))
self.g.add_node(2, (2, 3))
self.g.add_node(3, (6, 3))
self.g.add_node(4, (6, 7))
self.g.add_node(5, (7, 4))
self.g.add_node(6, (5, 3))
def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (1, 1))
self.g.add_node(2, (2, 3))
self.g.add_node(3, (6, 3))
self.g.add_node(4, (6, 7))
self.g.add_node(5, (7, 4))
self.g.add_node(6, (5, 3))
self.population = initial_population(self.g, 4)
class AlgorithmTests(unittest.TestCase):
def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (1, 1))
self.g.add_node(2, (2, 3))
self.g.add_node(3, (6, 3))
self.g.add_node(4, (6, 7))
self.g.add_node(5, (7, 4))
self.g.add_node(6, (5, 3))
# def test_random_search(self):
# rs = random_search(self.g, 10)
# self.assertEqual(rs[0], (4, 2, 1))
# self.assertEqual(rs[1], 6.23606797749979)
def test_two_opt_neighbourhood(self):
small_tour = (1, 2, 3)
two_opt = two_opt_neighbourhood(small_tour)
self.assertEqual(sorted(two_opt),
sorted([(2, 1, 3), (1, 3, 2), (3, 2, 1)]))
def test_best_neighbour(self):
tour = (1, 4, 2, 5, 3, 6)
self.assertEqual(best_neighbourhood(self.g, tour),
((1, 2, 4, 5, 3, 6), 17.941549404533227))
class GraphTests(unittest.TestCase):
def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (0, 0))
self.g.add_node(2, (0, 1))
self.g.add_node(3, (1, 1))
self.g.add_node(4, (1, 0))
def test_euclidean_distance(self):
self.assertEqual(self.g.euclidean_distance(1, 2), 1)
self.assertEqual(self.g.euclidean_distance(2, 3), 1)
self.assertEqual(self.g.euclidean_distance(3, 4), 1)
def test_tour_cost(self):
a = self.g.euclidean_distance(1, 2)
b = self.g.euclidean_distance(2, 3)
c = self.g.euclidean_distance(3, 4)
d = self.g.euclidean_distance(3, 4)
cost = a + b + c + d
self.assertEqual(self.g.get_tour_cost([1, 2, 3, 4]), cost)
class GeneticAlgorithmTests(unittest.TestCase):
def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (1, 1))
self.g.add_node(2, (2, 3))
self.g.add_node(3, (6, 3))
self.g.add_node(4, (6, 7))
self.g.add_node(5, (7, 4))
self.g.add_node(6, (5, 3))
self.population = initial_population(self.g, 4)
def test_initial_population(self):
self.assertEqual(len(self.population), 4)
def test_selection(self):
ranked_tours = rank_tours(self.g, self.population)
sel = selection(rank_tours, 2)
self.assertEqual(len(selection), 2)
from graph import CompleteGraph
from algorithms import (best_neighbourhood, random_search, swap_elements,
two_opt_neighbourhood, local_search)
from utils import from_csv
data = from_csv("ulysses16")
graph = CompleteGraph()
# Add information from ulysses16 to the CompleteGraph
for item in data:
name = item[0]
coordinate = (item[1], item[2])
graph.add_node(name, coordinate)
def testing_random_search():
limit = 20
print("Random Search Algorithm Example ({} seconds)".format(limit))
shortest = random_search(graph, limit)
print("Shortest Tour: {}".format(shortest[0]))
print("Cost: {}".format(shortest[1]))
def testing_two_opt_neighbourhood():
tour = graph.random_tour()
print("2-Opt Neighbourhood Example")
print("Tour: {}".format(tour))
for opt_tour in two_opt_neighbourhood(tour):
print(opt_tour)
from graph import CompleteGraph from weight_matrices import * c_graph = CompleteGraph(weight_matrices_array[1]) print(*c_graph.prim()) print(*c_graph.kruskal())
def setUp(self):
self.g = CompleteGraph()
self.g.add_node(1, (0, 0))
self.g.add_node(2, (0, 1))
self.g.add_node(3, (1, 1))
self.g.add_node(4, (1, 0))