def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
controller.title('Genetic Algorithm')
canvas = tk.Canvas(self, width=canvas_width, height=canvas_height)
canvas.pack(expand=tk.YES, fill=tk.BOTH, padx=20, pady=20)
population = search.init_population(max_population, gene_pool, len(target))
for generation in range(ngen):
population = [search.mutate(search.recombine(*search.select(2, population, fitness_fn)), gene_pool, mutation_rate) for i in range(len(population))]
current_best = ''.join(argmax(population, key=fitness_fn))
members = [''.join(x) for x in population][:48]
canvas.delete('all')
canvas.create_text(canvas_width / 2, 40, fill=p_blue, font='Consolas 46 bold', text=current_best)
for i in range(len(members) // 3):
canvas.create_text((canvas_width * .175), (canvas_height * .25 + (25 * i)), fill=p_blue, font='Consolas 16', text=members[3 * i])
canvas.create_text((canvas_width * .500), (canvas_height * .25 + (25 * i)), fill=p_blue, font='Consolas 16', text=members[3 * i + 1])
canvas.create_text((canvas_width * .825), (canvas_height * .25 + (25 * i)), fill=p_blue, font='Consolas 16', text=members[3 * i + 2])
canvas.create_text((canvas_width * .5), (canvas_height * 0.95), fill=p_blue, font='Consolas 18 bold', text=f'Generation {generation}')
canvas.update()
fittest_individual = search.fitness_threshold(fitness_fn, f_thres, population)
if fittest_individual:
break
def raise_frame(frame, init=False, update_target=False, target_entry=None, f_thres_slider=None): frame.tkraise() global target if update_target and target_entry is not None: target = target_entry.get() f_thres_slider.config(to=len(target)) if init: population = search.init_population(max_population, gene_pool, len(target)) genetic_algorithm_stepwise(population)
def main():
population = search.init_population(max_population, gene_pool, len(target))
mating_pool = []
for _ in range(1000):
new_population = [
search.mutate(
search.recombine(
*select(2, population, fitness_fn, mating_pool)),
gene_pool, mutation_rate)
]
fittest_individual = argmax(new_population, key=fitness_fn)
current_best = ''.join(fittest_individual)
if current_best == target:
return
print(current_best)
population = new_population
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
controller.title('Genetic Algorithm')
w = tk.Canvas(self, width=800, height=600)
w.pack(expand=tk.YES, fill=tk.BOTH)
# var = tk.StringVar()
# var.set('')
# label = tk.Label(self, textvariable=var)
# label.pack()
population = search.init_population(max_population, gene_pool,
len(target))
for i in range(100):
print(i)
# time.sleep(0.2)
population = [
search.mutate(
search.recombine(
*search.select(2, population, fitness_fn)), gene_pool,
mutation_rate) for i in range(len(population))
]
current_best = ''.join(argmax(population, key=fitness_fn))
w.create_text(100,
10,
fill='darkblue',
font='Times 20 italic bold',
text=current_best)
w.update()
# var.set(current_best)
# parent.update_idletasks()
# checks for completion
fittest_individual = search.fitness_threshold(
fitness_fn, f_thres, population)
if fittest_individual:
break
# v.set(current_best)
# self.update_idletasks()
# label.configure(text=current_best)
# label.update()
# solution, generations = genetic_algorithm_stepwise(self, population, fitness_fn, gene_pool, f_thres, ngen, mutation_rate)
controller.mainloop()
def main(): population = search.init_population(max_population, gene_pool, len(target)) solution, iterations = genetic_algorithm_stepwise(population, fitness_fn, f_thres=len(target), gene_pool=gene_pool, pmut=mutation_rate)
def main(): population = search.init_population(max_population, gene_pool, len(target)) solution = game_loop(population, fitness_fn, gene_pool=gene_pool, f_thres=f_thres, pmut=mutation_rate, ngen=ngen)
x2s[1:] = xs[:length - 1]
x2s[0] = xs[length - 1]
y2s[1:] = ys[:-1]
y2s[0] = ys[-1]
value = np.exp(
-np.log(np.sum(np.sqrt((x2s - xs)**2 +
(y2s - ys)**2)) + 0.00001) * 5 + 35)
return value
city_num = 15
cities = generate_random_cities(city_num=city_num, length=100)
population = init_population(pop_number=1000,
gene_pool=None,
state_length=city_num)
path, history = genetic_algorithm(population,
fitness_fn,
gene_pool=[0, 1],
f_thres=None,
ngen=200,
pmut=0.1)
print(path)
xs = []
ys = []
for i in range(len(path)):
xs.append(cities[path[i]][0])