def calc_cost(solution):
cost = tsplib.calc_cost(solution, distances)
sol_pens = penalties[solution[:-1], solution[1:]]
gls_cost = (cost +
alpha *
(best_cost/count+1) *
(sol_pens.sum() + penalties[solution[-1], solution[0]]))
return gls_cost
def nearest_tour(distances, count, randomize=True):
available = range(1, count)
if randomize:
shuffle(available)
tour = [0]
while available:
nearest = numpy.argmin(distances[tour[-1]][available])
tour.append(available[nearest])
available.pop(nearest)
solution = numpy.array(tour, dtype=numpy.int)
cost = tsplib.calc_cost(solution, distances)
return solution, cost
def main(argv):
alpha = float(argv[1])
coords, distances = tsplib.load(argv[2])
count = len(distances)
solution, best_cost = nearest_tour(distances, len(distances))
plot = plotting.TspPlot(coords, draw_guess=True)
print best_cost
penalties = numpy.copy(distances)
penalties[:] = 0.0
def calc_cost(solution):
cost = tsplib.calc_cost(solution, distances)
sol_pens = penalties[solution[:-1], solution[1:]]
gls_cost = (cost +
alpha *
(best_cost/count+1) *
(sol_pens.sum() + penalties[solution[-1], solution[0]]))
return gls_cost
def penalise(solution):
sol_dists = distances[solution[0:-1], solution[1:]]
sol_dists = numpy.append(sol_dists, (distances[solution[0], solution[-1]],))
sol_pens = penalties[solution[0:-1], solution[1:]]
sol_pens = numpy.append(sol_pens, (penalties[solution[0], solution[-1]],))
utility = sol_dists/(sol_pens+1)
index_a = numpy.argmax(utility)
index_b = index_a+1 if index_a != count-1 else 0
a = solution[index_a]
b = solution[index_b]
penalties[a,b] += 1.0
penalties[b,a] += 1.0
def redraw_guess(solution, cost):
cost = tsplib.calc_cost(solution, distances) # ugh
plot.redraw_guess(solution, cost)
gls_cost = cost = best_cost
best_solution = numpy.copy(solution)
plot.redraw_best(best_solution, cost)
while 1:
solution, gls_cost = two_opt(solution, gls_cost, len(distances), calc_cost, redraw_guess)
cost = tsplib.calc_cost(solution, distances)
if cost < best_cost:
print cost
best_cost = cost
best_solution[:] = solution
plot.redraw_best(best_solution, cost)
penalise(solution)
def calc_cost(solution):
return tsplib.calc_cost(solution, distances)
def redraw_guess(solution, cost):
cost = tsplib.calc_cost(solution, distances) # ugh
plot.redraw_guess(solution, cost)
def calc_cost(solution):
return tsplib.calc_cost(solution, distances)
