Skip to content

Instantly share code, notes, and snippets.

@howCodeORG
Created May 12, 2018 01:35
Show Gist options
  • Save howCodeORG/5a141753b1c7b1abe6284d8952bf612e to your computer and use it in GitHub Desktop.
Save howCodeORG/5a141753b1c7b1abe6284d8952bf612e to your computer and use it in GitHub Desktop.
howCode's Simple Genetic Algorithm in Python
import random
population = 200
generations = 0
mutation = 0.01
alphabet = "abcdefghijklmnopqrstuvwxyz! "
target = "subscribe to howcode!"
output = ""
data = []
pool = []
score_range = []
class Item:
def __init__(self, data, target):
self.target = target
self.data = data
self.score = self.get_score()
def get_score(self):
score = 0
for i in range(len(self.data)):
if self.data[i] == self.target[i]:
score += 1
return score / len(self.data)
def __str__(self):
return 'String: ' + ''.join(self.data) + ', Score: ' + str(self.score)
# SETUP
for i in range(population):
data.append(Item([random.choice(alphabet) for item in [0] * len(target)], target))
while output != target:
pool = []
# SELECTION
for item in data:
if item != 0:
for i in range(int(item.score * 100)):
pool.append(item)
# REPEAT
# PICK 2 PARENTS
# CROSSOVER
# MUTATION
# ADD NEW CHILD TO POPULATION
data = []
while len(data) < population:
parentA = pool[random.randint(0,len(pool)-1)]
parentB = pool[random.randint(0,len(pool)-1)]
parentAScore = int(parentA.score / (parentA.score + parentB.score) * 100)
parentBScore = int(parentB.score / (parentA.score + parentB.score) * 100)
childData = []
for i in range(len(target)):
choice_list = [parentA.data[i]] * int(parentAScore) + [parentB.data[i]] * int(parentBScore)
childData.append(random.choice(choice_list))
for i in range(len(childData)):
m = mutation * 100
r = random.randint(0,100/m)
if r == 0:
childData[i] = random.choice(alphabet)
child = Item(childData, target)
data.append(child)
output = "".join(child.data)
if output == target:
break
best = None
for i in range(len(data)):
if best == None:
best = data[i]
elif data[i].score > best.score:
best = data[i]
print(best)
generations += 1
print("Generation: " + str(generations))
@felipevlima
Copy link

`import random

population = 200
poolLength = 20
generations = 0
mutation = 0.01

alphabet = "abcdefghijklmnopqrstuvwxyz!?ABCDEFGHIJKLMNOPQRSTUVWXYZáéíóúÁÉÍÓÚâêîôûÂÊÎÔÛçÇ.,/|=+-_{}[]()&@#$%*:;ÃÕãõ1234567890~˜'` "
target = "jaohuikjhenkujn``EDPoLHDSOAHUIEKJ"
output = ""
data = []
pool = []
score_range = []

class Item:
    def __init__(self, data, target):
        self.target = target
        self.data = data
        self.score = self.get_score()

    def get_score(self):
        score = 0
        for i in range(len(self.data)):
            if self.data[i] == self.target[i]:
                score += 1
        return score / len(self.data)

    def __str__(self):
        return 'String: ' + ''.join(self.data) + ', Score: ' + str(self.score)

# SETUP
for i in range(population):
    randomString = ''
    for j in range(len(target)):
        randomString += random.choice(alphabet)
    data.append(Item(randomString, target))

while output != target:

    pool = []

    # SELECTION
    sortedData = sorted(data, key=lambda item: item.score, reverse=True)
    pool = sortedData[0: poolLength]

    # REPEAT
    # PICK 2 PARENTS
    # CROSSOVER
    # MUTATION
    # ADD NEW CHILD TO POPULATION

    data = []
    while len(data) < population:
        parentA = pool[random.randint(0,len(pool)-1)]
        parentB = pool[random.randint(0,len(pool)-1)]

        parentAScore = parentA.score / (parentA.score + parentB.score)

        childData = []
        for i in range(len(target)):
            probability = random.uniform(0,1)
            if probability <= parentAScore:
                childData.append(parentA.data[i])
            else:
                childData.append(parentB.data[i])

        for i in range(len(childData)):
            m = mutation * 100
            r = random.randint(0,100/m)
            if r == 0:
                childData[i] = random.choice(alphabet)

        child = Item(childData, target)
        data.append(child)
        output = "".join(child.data)
        if output == target:
            break
    best = None
    for i in range(len(data)):
        if best == None:
            best = data[i]
        elif data[i].score > best.score:
            best = data[i]
    print(best)
    generations += 1
    print("Generation: " + str(generations))`

+EDITED
+MORE LETTES

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment