genetic.py

import random
import time

chars = []
for i in range(32, 127):
    chars.append(chr(i))

def randomString():
    """Generate a random string of fixed length """
    return ''.join(random.choice(chars) for i in range(28))

def initialise_population(population_size:int):
    population = []
    for i in range(0, population_size):
        population.append(randomString())
    return population

def fitness_eval(member:str):
    string = 'methinks it is like a weasel'
    fitness = 0
    split = list(string)
    for i in range(0, 28):
        if member[i] == split[i]:
            fitness = fitness+1
    return fitness

def mutation(parent: str, mutation_rate=70):
    parent = list(parent)
    child = [0] * 28
    for i in range(0, 28):
        p = random.uniform(0, 1)
        if (p) < (1/mutation_rate):
            child[i] = random.choice(chars)
        else:
            child[i] = parent[i]
    return child

def crossover(parentA,parentB):
    child = [0] * 28
    for i in range(0,28):
        if random.uniform(0,1) < 0.5:
            child[i] = parentA[i]
        else:
            child[i] = parentB[i]
    return child

def choose_parent(pop):
    A = random.choice(pop)
    B = random.choice(pop)
    if fitness_eval(A) > fitness_eval(B):
        parent = A
    else:
        parent = B
    return parent

def genetic_algorithm_crossover(population_size):
    pop = initialise_population(population_size)
    fitness,counter = 0,0

    while(fitness<28):
        print(counter)
        A = random.choice(pop)
        B = random.choice(pop)

        if fitness_eval(A) > fitness_eval(B):
            parent1 = A
        else:
            parent1 = B

        A = random.choice(pop)
        B = random.choice(pop)

        if fitness_eval(A) > fitness_eval(B):
            parent2 = A
        else:
            parent2 = B

        res = crossover(parent1, parent2) #create a child
        child = mutation(res)

        A = random.choice(pop)
        B = random.choice(pop)
        iA = pop.index(A)
        iB = pop.index(B)

        if (fitness_eval(A) > fitness_eval(B)):  # if A is better than B
            pop[iB] = child  # B is child
        else:
            pop[iA] = child

        if fitness_eval(child) > fitness:
            fitness = fitness_eval(child)
            print(fitness,child)
        counter+=1

genetic_algorithm_crossover(500)