def minimize():
global elite
population = [gf.get_individual(utility.n_features) for i in range(go.population_size)]
for i in range(n_iter):
population_new = []
if elite:
population_new.append(max(population))
while len(population_new) < go.population_size:
parents = random.sample(population, 2)
child = parents[0] + parents[1]
population_new.append(~child)
population = population_new
if 0 == i % 500:
print(max(population))
return max(population)
def minimize(n_iter_trace = False):
global f_eps
population = [gf.get_individual(utility.n_features) for i in range(go.population_size)]
for i in range(n_iter):
tournament = random.sample(population, go.tournament_size)
worst = min(tournament)
tournament.remove(worst)
population.remove(worst)
parents = random.sample(tournament, 2)
child = parents[0] + parents[1]
new_individual = ~child
if new_individual > max(population):
info(new_individual, i) if printed else None
if abs(new_individual.fitness) < f_eps : return new_individual if not n_iter_trace else new_individual, i
population.append(new_individual)
if n_iter_trace: return max(population), n_iter
return max(population)
