def simulate_evolution(self, pop_count=100, gen_count=150000, save_folder="test/"):
# initializes population
tot_fitness = 0 # total fitness of generation
population = []
pop_weight = []
most_fit_org = circle_org()
population.append(most_fit_org)
pop_weight.append(most_fit_org.fit_score)
tot_fitness += most_fit_org.fit_score
for c in range(pop_count - 1):
new_org = circle_org()
population.append(new_org)
pop_weight.append(new_org.fit_score)
tot_fitness += new_org.fit_score
if new_org.fit_score > most_fit_org.fit_score:
most_fit_org = new_org
# begins loop for evolution
cond = True # TODO: stop loop if certain fitscore is met
for x in range(0, gen_count):
# saving most fit
most_fit_org.save_image("test/" + str(x) + " " + str(most_fit_org.fit_score) + ".png")
print(most_fit_org.fit_score)
most_fit_org.fit_score = 0 # TODO: DIRTY, FIX THIS
temp_population = [] # temporary population buff
new_pop_count = 0
while new_pop_count < pop_count:
# 2 new individuals
p1 = weighted_choice(population, pop_weight)
p2 = weighted_choice(population, pop_weight)
# if yes_no(circ_org_handler.CROSS_RATE):
# select 2 parents
kids = self.gen_new_org(p1, p2)
for kid in kids:
if kid.fit_score > p1.fit_score or kid.fit_score > p2.fit_score:
temp_population.append(kid)
new_pop_count += 1
if kid.fit_score >= most_fit_org.fit_score:
most_fit_org = kid
population = temp_population
def gen_new_org(self, org1, org2):
new_org_dna = self.__crossover__(org1.DNA, org2.DNA)
for dna in new_org_dna:
self.__mutate__(dna)
# print (new_org_dna[1].bin)
return (circle_org(new_org_dna[0], False), circle_org(new_org_dna[1], False))
