self.route, self.cur_cost = self.greedy_solution()
while self.temperature >= self.stopping_temperature and self.iteration < self.stopping_iter:
guess = list(self.route)
left_index = random.randint(2, self.num_cities - 1)
right_index = random.randint(0, self.num_cities - left_index)
guess[right_index: (right_index + left_index)] = reversed(guess[right_index: (right_index + left_index)])
self.accept(guess)
self.temperature *= self.alpha
self.iteration += 1
self.progress.append(self.cur_cost)
print("Best fitness obtained: ", self.best_fitness)
def visualize_routes(self):
visualize_tsp('simulated annealing TSP', self.route)
def plot_learning(self):
fig = plt.figure(1)
plt.plot([i for i in range(len(self.progress))], self.progress)
plt.ylabel("Distance")
plt.xlabel("Iterations")
plt.show(block=False)
if __name__ == "__main__":
cities = read_cities(64)
sa = SimAnneal(cities, stopping_iter=15000)
sa.run()
sa.plot_learning()
sa.visualize_routes()
gbest[swap[0]], gbest[swap[1]] = gbest[swap[1]], gbest[
swap[0]]
particle.velocity = temp_velocity
for swap in temp_velocity:
if random.random() <= swap[2]:
new_route[swap[0]], new_route[swap[1]] = \
new_route[swap[1]], new_route[swap[0]]
particle.route = new_route
particle.update_costs_and_pbest()
if __name__ == "__main__":
cities = read_cities(448)
pso = PSO(iterations=3200,
population_size=300,
pbest_probability=0.9,
gbest_probability=0.02,
cities=cities)
pso.run()
print(f'cost: {pso.gbest.pbest_cost}\t| gbest: {pso.gbest.pbest}')
x_list, y_list = [], []
for city in pso.gbest.pbest:
x_list.append(city.x)
y_list.append(city.y)
x_list.append(pso.gbest.pbest[0].x)
y_list.append(pso.gbest.pbest[0].y)
fig = plt.figure(1)
gbest[swap[0]], gbest[swap[1]] = gbest[swap[1]], gbest[
swap[0]]
particle.velocity = temp_velocity
for swap in temp_velocity:
if random.random() <= swap[2]:
new_route[swap[0]], new_route[swap[1]] = \
new_route[swap[1]], new_route[swap[0]]
particle.route = new_route
particle.update_costs_and_pbest()
if __name__ == "__main__":
cities = read_cities(29)
pso = PSO(iterations=1200,
population_size=300,
pbest_probability=0.9,
gbest_probability=0.02,
cities=cities)
pso.run()
print(f'cost: {pso.gbest.pbest_cost}\t| gbest: {pso.gbest.pbest}')
x_list, y_list = [], []
for city in pso.gbest.pbest:
x_list.append(city.x)
y_list.append(city.y)
x_list.append(pso.gbest.pbest[0].x)
y_list.append(pso.gbest.pbest[0].y)
fig = plt.figure(1)
import itertools
import matplotlib.pyplot as plt
from util import City, read_cities, write_cities_and_return_them, generate_cities, path_cost
def solve_tsp_dynamic(cities):
distance_matrix = [[x.distance(y) for y in cities] for x in cities]
cities_a = {(frozenset([0, idx + 1]), idx + 1): (dist, [0, idx + 1]) for idx, dist in
enumerate(distance_matrix[0][1:])}
for m in range(2, len(cities)):
cities_b = {}
for cities_set in [frozenset(C) | {0} for C in itertools.combinations(range(1, len(cities)), m)]:
for j in cities_set - {0}:
cities_b[(cities_set, j)] = min([(cities_a[(cities_set - {j}, k)][0] + distance_matrix[k][j],
cities_a[(cities_set - {j}, k)][1] + [j])
for k in cities_set if k != 0 and k != j])
cities_a = cities_b
res = min([(cities_a[d][0] + distance_matrix[0][d[1]], cities_a[d][1]) for d in iter(cities_a)])
return res[1]
if __name__ == "__main__":
cities = read_cities(16)
g = solve_tsp_dynamic(cities)
sol = [cities[gi] for gi in g]
print(path_cost(sol))
plt.show(block=True)