def test_bad_given_thresholds_not_in_valid_range(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
bad_range_thresholds = [THRESHOLD_MAX * 2] * total_bits
with pytest.raises(InvalidSpecError):
make_real_mux_env(num_address_bits=num_address_bits,
thresholds=bad_range_thresholds)
def test_bad_given_thresholds_not_float(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
non_float_thresholds = [int(THRESHOLD_MIN)] * total_bits
with pytest.raises(InvalidSpecError):
make_real_mux_env(num_address_bits=num_address_bits,
thresholds=non_float_thresholds)
def test_bad_given_thresholds_incorrect_len_2d_array(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
bad_len_thresholds = [[0.5] * total_bits, [0.5] * total_bits]
with pytest.raises(InvalidSpecError):
make_real_mux_env(num_address_bits=num_address_bits,
thresholds=bad_len_thresholds)
def test_labels_values(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_bits
mux = make_real_mux_env(num_address_bits=num_address_bits,
thresholds=thresholds)
assert np.all((mux.labels == 0) | (mux.labels == 1))
def test_data_values(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_bits
mux = make_real_mux_env(num_address_bits=num_address_bits,
thresholds=thresholds)
assert np.all(0.0 <= mux.data) and np.all(mux.data < 1.0)
def test_labels_dtype(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_bits
mux = make_real_mux_env(num_address_bits=num_address_bits,
thresholds=thresholds)
for elem in mux.labels.flatten():
assert isinstance(elem, np.integer)
def test_data_dtype(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_bits
mux = make_real_mux_env(num_address_bits=num_address_bits,
thresholds=thresholds)
for elem in mux.data.flatten():
assert isinstance(elem, np.floating)
def test_obs_space_integrity(self):
num_address_bits = 1
total_feature_dims = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_feature_dims
mux = make_real_mux_env(thresholds=thresholds, num_address_bits=1)
assert len(mux.obs_space) == total_feature_dims
for dim in mux.obs_space:
assert dim.lower == 0.0
assert dim.upper == 1.0
def test_action_set_integrity(self):
num_address_bits = 1
total_bits = calc_total_bits(num_address_bits)
thresholds = [0.5] * total_bits
mux = make_real_mux_env(thresholds=thresholds, num_address_bits=1)
assert mux.action_set == {0, 1}
def test_bad_num_address_bits_negative(self):
with pytest.raises(InvalidSpecError):
make_real_mux_env(num_address_bits=-1, thresholds=[])
def test_sample_training_run(self):
# 6-mux
env = make_real_mux_env(thresholds=[0.5] * 6,
num_address_bits=2,
shuffle_dataset=True,
reward_correct=1000,
reward_incorrect=0,
num_samples=64)
rule_repr = CentreSpreadRuleRepr(env.obs_space)
alg_hyperparams = {
"N": 400,
"beta": 0.2,
"alpha": 0.1,
"epsilon_nought": 0.01,
"nu": 5,
"gamma": 0.71,
"theta_ga": 12,
"chi": 0.8,
"mu": 0.04,
"theta_del": 20,
"delta": 0.1,
"theta_sub": 20,
"p_wildcard": 0.33,
"prediction_I": 1e-3,
"epsilon_I": 1e-3,
"fitness_I": 1e-3,
"p_explore": 0.5,
"theta_mna": len(env.action_set),
"do_ga_subsumption": True,
"do_as_subsumption": True,
"seed": 0,
"m": 0.1,
"s_nought": 1.0
}
alg = make_canonical_xcs(env, rule_repr, alg_hyperparams)
monitor_items = [
MonitorItem("num_micros",
lambda experiment: experiment.population.num_micros),
MonitorItem("num_macros",
lambda experiment: experiment.population.num_macros),
MonitorItem(
"performance", lambda experiment: experiment.calc_performance(
strat="accuracy")),
MonitorItem(
"mean_error", lambda experiment: calc_summary_stat(
experiment.population, "mean", "error")),
MonitorItem(
"max_fitness", lambda experiment: calc_summary_stat(
experiment.population, "max", "fitness")),
MonitorItem(
"deletion", lambda experiment: experiment.population.
operations_record["deletion"]),
MonitorItem(
"covering", lambda experiment: experiment.population.
operations_record["covering"]),
MonitorItem(
"as_subsumption", lambda experiment: experiment.population.
operations_record["as_subsumption"]),
MonitorItem(
"ga_subsumption", lambda experiment: experiment.population.
operations_record["ga_subsumption"]),
MonitorItem(
"discovery", lambda experiment: experiment.population.
operations_record["discovery"]),
MonitorItem(
"absorption", lambda experiment: experiment.population.
operations_record["absorption"])
]
experiment = Experiment(env,
alg,
num_training_epochs=10,
monitor_items=monitor_items)
experiment.run()
print(experiment._monitor.query())