optimising_rastrigin

Example Python program optimising_rastrigin.py

Modules

• import numpy as np
• import matplotlib.pyplot as plt
• from matplotlib import animation, rc

Methods

• def plot(surface, positions, best_fitness, best_position, iteration):
• def run(iteration):

Code

Python example

# Imports
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import animation, rc


# A way of plotting each step of the optimisation process.
def plot(surface, positions, best_fitness, best_position, iteration):
    plt.cla()

    # Plot the feature / error surface.
    plt.pcolormesh(surface[0], surface[1], surface[2])

    # Plot all of the genepool.
    x, y = zip(*positions)
    plt.scatter(x, y, 1, 'k', edgecolors='face')

    # Plot the best gene.
    plt.scatter(best_position[0], best_position[1], 20, 'w', edgecolors='face')

    # Add the title to the plot.
    title = "iteration {}, fitness {}".format(iteration, best_fitness)
    plt.title(title)


# Each time this function is ran, the optimisation process is furthered a step.
def run(iteration):

    # Get the rastrigin error surface.
    surface = function.get_surface()

    # Get the gene pool.
    positions = ga.get_solutions()

    # Evaluate the fitnesses on the rastrigin surface
    fitnesses = [function.evaluate(pos) for pos in positions]

    # Inform the GA of the genepool's performance
    ga.set_fitnesses(fitnesses)

    # Get the best gene
    best_position, best_fitness = ga.get_best()

    # Plot the optimsation
    plot(surface, positions, best_fitness, best_position, iteration)


# Genetic Algorithm parameters 
elitism = 0.1
population_size = 500
mutation_rate = 0.8
mutation_sigma = 0.1
mutation_decay = 0.999
mutation_limit = 0.01
amount_optimisation_steps = 250
dna_bounds = (-5.11, 5.11)
dna_start_position = [4.8, 4.8]

# Construct the test function
function = Rastrigin()

# Construct the GA
ga = GA(len(dna_start_position),
        dna_bounds,
        dna_start_position,
        elitism,
        population_size,
        mutation_rate,
        mutation_sigma,
        mutation_decay,
        mutation_limit)

# Create the matplotlib figure
fig = plt.figure(figsize=(10, 10))

# See this on making gifs with matplotlib:
# https://matplotlib.org/api/animation_api.html
anim = animation.FuncAnimation(fig, run, frames=amount_optimisation_steps)
anim.save('rastrigin3_ga.mp4', fps=15)

Useful Links

• Articles: https://python-commandments.org/
• Python shell: https://bsdnerds.org/learn-python/
• Tutorial: https://pythonprogramminglanguage.com/