Genetic Algorithm (GA)¶

In [1]:

from metazoo.bio.evolutionary import GeneticAlgorithm
from metazoo.bio.evolutionary.operators import selection, mutation, crossover
from metazoo.gym.mono import Function
from metazoo.bio.utils import encoding

from metazoo.bio.evolutionary import GeneticAlgorithm from metazoo.bio.evolutionary.operators import selection, mutation, crossover from metazoo.gym.mono import Function from metazoo.bio.utils import encoding

In [2]:

available = Function().available_functions
print(available)

available = Function().available_functions print(available)

['Rastrigin', 'Ackley', 'Sphere', 'Rosenbrock', 'Beale', 'GoldsteinPrice', 'Booth', 'Bukin', 'Matyas', 'Levi_N13', 'Griewank', 'Himmelblau', 'ThreeHumpCamel', 'Easom', 'Cross_In_Tray', 'EggHolder', 'HolderTable', 'McCormick', 'Schaffer_N2', 'StyblinskiTang', 'Shekel']

In [3]:

fitness_function = Function('Ackley', reverse=False)

fitness_function = Function('Ackley', reverse=False)

In [4]:

fitness_function.bounds

fitness_function.bounds

Out[4]:

[(-5, 5), (-5, 5)]

In [5]:

#fitness_function.plot(bounds=fitness_function.bounds[0], dim=2, num_points=100, mode='surface')
#fitness_function.plot(bounds=fitness_function.bounds[0], dim=2, num_points=100, mode='contour')

#fitness_function.plot(bounds=fitness_function.bounds[0], dim=2, num_points=100, mode='surface') #fitness_function.plot(bounds=fitness_function.bounds[0], dim=2, num_points=100, mode='contour')

In [11]:

type = 'binary'

if type == 'binary':
    mutation_function = mutation.flip_bit
    encoder = encoding.Binary(precision=3, bounds=fitness_function.bounds)
else:
    mutation_function = mutation.gaussian
    encoder = encoding.Real(bounds=fitness_function.bounds)

ga = GeneticAlgorithm(
    fitness_function=fitness_function,
    crossover_function=crossover.onepoint,
    mutation_function=mutation_function,
    selection_function=selection.tournament,
    population_size=200,
    mutation_rate=0.01,
    crossover_rate=0.7,
    encoder=encoder,
    minimize=True,
)

#ga.summary()

type = 'binary' if type == 'binary': mutation_function = mutation.flip_bit encoder = encoding.Binary(precision=3, bounds=fitness_function.bounds) else: mutation_function = mutation.gaussian encoder = encoding.Real(bounds=fitness_function.bounds) ga = GeneticAlgorithm( fitness_function=fitness_function, crossover_function=crossover.onepoint, mutation_function=mutation_function, selection_function=selection.tournament, population_size=200, mutation_rate=0.01, crossover_rate=0.7, encoder=encoder, minimize=True, ) #ga.summary()

In [12]:

ga.run(generations=500)

ga.run(generations=500)

Output()

In [13]:

ga.best_fitness

ga.best_fitness

Out[13]:

0.001225738155231415

In [14]:

best = ga.best_individual.reshape(1, -1)
print(f'Best Individual: {best}')

best = ga.best_individual.reshape(1, -1) print(f'Best Individual: {best}')

Best Individual: [[-0.00030519  0.00030519]]

In [15]:

x = best[0]
result = fitness_function(x)
print(result)

x = best[0] result = fitness_function(x) print(result)

0.001225738155231415

In [16]:

fig1 = fitness_function.plot(bounds=fitness_function.bounds, dim=2, num_points=100, mode='surface', population=best)
fig2 = fitness_function.plot(bounds=fitness_function.bounds, dim=2, num_points=100, mode='contour', population=best)

fig1 = fitness_function.plot(bounds=fitness_function.bounds, dim=2, num_points=100, mode='surface', population=best) fig2 = fitness_function.plot(bounds=fitness_function.bounds, dim=2, num_points=100, mode='contour', population=best)

In [17]:

fig1.show()
fig2.show()

fig1.show() fig2.show()

In [18]:

ga.fitness_plot()

ga.fitness_plot()