Genetic Hello World

import random
from collections import namedtuple

def evolve(initialise_population_f, fitness_f, breed_f, num_generations=100):
    """
    initialise_population: Returns an initial population represented as a list
    fitness_f            : Fitness function
    breed_f              : Takes two elements and breeds them, possibly introducing mutations
    num_generations      : Number of generations to run the evolver for
    """

    Individual = namedtuple("Individual", ["fitness", "gene"])

    population = [Individual(fitness_f(x), x) for x in initialise_population_f()]
    pop_size = len(population)

    def rand_index():
        return max(pop_size - 1, int(pop_size * random.triangular(0, 1, 1)))

    for generation in range(num_generations):
        new_population = []
        population.sort()

        for i in range(pop_size):
            index1 = rand_index()
            index2 = rand_index()

            new_indiv = breed_f(population[index1].gene, population[index2].gene)
            new_population.append(Individual(fitness_f(new_indiv), new_indiv))

        population = new_population

    population.sort()
    return population[-1]

alphabet = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ, !"

def rand_string(n, alphabet):
        return "".join([random.choice(alphabet) for i in range(n)])

def initialise_population():
    return [rand_string(13, alphabet) for i in range(1000)]

def fitness_function(string):
    return sum((1 if string[i] == "Hello, World!"[i] else 0) for i in range(len(string)))

def breed(string1, string2):
    global alphabet

    index = random.randint(0, len(string1) - 1)
    child = string1[:index] + string2[index:]

    if random.random() < 0.1:
        mutation_index = random.randint(0, len(child) - 1)
        child = child[:mutation_index] + random.choice(alphabet) + child[mutation_index+1:]

    return child