def test_assembly_factory(self):
def make_g3(fval):
g = G3()
g.fval = fval
return g
@pyvotune.sparse_input
@pyvotune.pfloat(range=(-3, 72))
@pyvotune.factory(make_g3)
class G3:
def __init__(self):
pass
gen = pyvotune.Generate(
gene_pool=[G1, G3],
max_length=2, noop_frequency=0.)
genome = pyvotune.Genome("testid")
genome.add_gene([2, True], G1)
genome.add_gene([2.5], G2)
self.assertTrue(genome.assemble())
individual = genome.assembled
self.assertEqual(individual[0].ival, 2)
self.assertEqual(individual[0].bval, True)
self.assertEqual(individual[1].fval, 2.5)
def test_params(self):
gen = pyvotune.Generate(
gene_pool=[G1, G2],
max_length=2, noop_frequency=0.)
for i in range(10):
genome = gen.generate()
self.assertTrue(genome.assemble())
def test_gen(self):
gen = pyvotune.Generate(
gene_pool=[G1, G2],
max_length=2, noop_frequency=0.)
for i in range(10):
genome = gen.generate()
self.assertTrue(genome)
self.assertEqual(len(genome), 5)
self.assertEqual(genome[2][1], G1)
self.assertEqual(genome[4][1], G2)
gen = pyvotune.Generate(
gene_pool=[G1, G2],
max_length=2, noop_frequency=1.)
for i in range(10):
genome2 = gen.generate()
self.assertFalse(genome2)
def test_assembly(self):
gen = pyvotune.Generate(
gene_pool=[G1, G2],
max_length=2, noop_frequency=0.)
genome = pyvotune.Genome("testid")
genome.add_gene([2, True], G1)
genome.add_gene([2.5], G2)
self.assertTrue(genome.assemble())
individual = genome.assembled
self.assertEqual(individual[0].ival, 2)
self.assertEqual(individual[0].bval, True)
self.assertEqual(individual[1].fval, 2.5)
def test_random_reset(self):
@pyvotune.pint(
range=(0, 5))
class P1:
def __init__(self, intparam):
self.intparam = intparam
@pyvotune.pint(
range=(5, 10))
class P2:
def __init__(self, intparam):
self.intparam = intparam
gen = pyvotune.Generate(
gene_pool=[P1, P2],
max_length=2, noop_frequency=0.)
g1 = pyvotune.Genome(get_id())
g1.add_gene([3], P1)
g1.add_gene([7], P2)
self.assertTrue(g1.assemble())
r = random.Random()
prev_val = g1[0][1]
some_diff = False
for i in range(100):
(g2,) = pyvotune.variators.random_reset_mutation(
r, [g1], {
'mutation_rate': 1.0,
'pyvotune_generator': gen
})
new_val = g2[0][1]
if g1[0] != g2[0] or g1[1] != g2[1]:
some_diff = True
self.assertTrue(some_diff)
pyvotune.set_debug(True)
# Dummy data
n_features = 28 * 28
rng = random.Random()
#################################
# Initialize PyvoTune Generator #
#################################
gen = pyvotune.Generate(
initial_state={
'sparse': False
},
gene_pool=pyvotune.sklearn.get_classifiers(n_features, rng) +
pyvotune.sklearn.get_decomposers(n_features, rng) +
pyvotune.sklearn.get_image_features(n_features, rng) +
pyvotune.sklearn.get_preprocessors(n_features, rng),
max_length=4,
noop_frequency=0.2,
rng=rng)
args = {
'crossover_rate': 0.5,
'mutation_rate': 0.3,
'pyvotune_generator': gen
}
# Use PyvoTun variators
variators = [
pyvotune.variators.random_reset_mutation,
log.error("Need path to model -v")
sys.exit(0)
classify_models(app_args.validate[0], app_args.classify[0])
sys.exit(1)
elif app_args.validate:
validate_models(app_args.validate[0])
sys.exit(1)
if not app_args.worker_mode:
#################################
# Initialize PyvoTune Generator #
#################################
gen = pyvotune.Generate(initial_state={'sparse': False},
gene_pool=gene_pool,
max_length=app_args.max_length,
noop_frequency=0.2,
rng=rng)
####################################
# Initialize Inspyred Genetic Algo #
####################################
ea = inspyred.ec.cEA(rng)
ea.logger = log
ea.terminator = [
#inspyred.ec.terminators.time_termination,
inspyred.ec.terminators.average_fitness_termination
]
ea.selector = inspyred.ec.selectors.fitness_proportionate_selection
ea.archiver = pyvotune.archivers.pickle_wrap_archiver
print "Actual Solution:" "E = m *", SPEED_OF_LIGHT, "*", SPEED_OF_LIGHT
print "Best Solution:", best_eq
print "Actual C:", SPEED_OF_LIGHT**2
print "Our C:", best_eq(1)
print "Diff:", abs(best_eq(1) - SPEED_OF_LIGHT**2)
print "Diff Pct:", round(
abs(best_eq(1) - SPEED_OF_LIGHT**2) / (SPEED_OF_LIGHT**2.) * 100, 2)
print "Fitness", best.fitness
print "MSE", sum_errs / samps
if __name__ == '__main__':
pyvotune.set_debug(False)
gen = pyvotune.Generate(gene_pool=[mass, someconst, oper],
max_length=10,
noop_frequency=0.2)
ea = inspyred.ec.GA(random.Random())
ea.terminator = [
inspyred.ec.terminators.time_termination,
inspyred.ec.terminators.average_fitness_termination
]
ea.observer = inspyred.ec.observers.stats_observer
ea.variator = [
pyvotune.variators.param_reset_mutation,
pyvotune.variators.scramble_mutation,
pyvotune.variators.n_point_crossover
]
# Split the dataset into training, testing and then validation parts
train_X, temp_X, train_y, temp_y = train_test_split(X, y, test_size=0.25)
test_X, validate_X, test_y, validate_y = train_test_split(temp_X,
temp_y,
test_size=0.5)
n_features = X.shape[1]
#################################
# Initialize PyvoTune Generator #
#################################
gen = pyvotune.Generate(
initial_state={'sparse': False},
gene_pool=pyvotune.sklearn.get_regressors(n_features) +
pyvotune.sklearn.get_decomposers(n_features) +
pyvotune.sklearn.get_preprocessors(n_features),
max_length=4,
noop_frequency=0.2)
####################################
# Initialize Inspyred Genetic Algo #
####################################
ea = inspyred.ec.GA(random.Random())
ea.terminator = [
inspyred.ec.terminators.time_termination,
inspyred.ec.terminators.average_fitness_termination
]
ea.observer = inspyred.ec.observers.stats_observer
