gistfile1.py

#! /usr/bin/env python

from random import choice, randint

def random_chromosome():
    return [randint(0, 1) for i in range(32)]

def calc_fitness(chromo):
    return sum(chromo)

def mate(chromo1, chromo2):
    return mutate(chromo1[:16] + chromo2[16:])

def mutate(chromo):
    for i in range(2):
        chromo[randint(0, 31)] = randint(0, 1)
    return chromo

def alphas(arrs):
    decorated = [(calc_fitness(chromo), chromo) for chromo in arrs]
    decorated.sort(reverse = True)
    return [chromo[1] for chromo in decorated][:10]

def is_spartan(chromo):
    return sum(chromo) == sum([1] * len(chromo))

if __name__ == '__main__':
    from time import asctime as time
    from sys import argv

    if len(argv) < 2:
        NUM_OFFSPRING = 20
    else:
        NUM_OFFSPRING = int(argv[1])

    SPARTAN = [1] * 32
    SEEDS = [random_chromosome() for i in range(NUM_OFFSPRING)]
    top10 = alphas(SEEDS)

    i = 0
    while not SPARTAN in top10:
        print("*** ALPHAS FROM GENERATION: #%i ***" % (i))
        offspring = []
        for j in range(NUM_OFFSPRING):
            offspring.append(mate(choice(top10), choice(top10)))

        top10 = alphas(offspring)
        print(top10)
        i += 1

• there is random.choice
• NUM_OFFSPRING must be converted to int
• sum([1] * len(chromo)) is the same as len(chromo)

• Yes, I should have used that.
• Woops, a hasty mistake, nice catch.
• You're right here too, but this was a deliberate action. What [1] * len(chromo) does is generate the template gene we are matching against. I felt this was more expressive, but I can see how it is confusing.

I've applied changes for the first two, thanks for the suggestions!