def _test_plot_data():
domain = Domain.make(["a"], ["x", "y"], [(0, 1), (0, 1)])
a, x, y = domain.get_symbols(["a", "x", "y"])
formula = a | (~a & (x <= y))
data = uniform(domain, 100)
labels = evaluate(domain, formula, data)
mpl.use('Agg')
plot_data(None, domain, (data, labels))
assert True
def select_active(
self, domain, data, labels, formula,
active_indices) -> Tuple[np.ndarray, np.ndarray, List[int]]:
active_set = set(active_indices)
learned_labels = evaluate(domain, formula, data)
differences = np.logical_xor(labels, learned_labels)
difference_set = set(np.where(differences)[0])
# print(active_set)
# print(difference_set)
# print(pretty_print(formula))
# for i in active_set & difference_set:
# print(i)
# print(pretty_print_instance(domain, data[i]))
# print(labels[i], learned_labels[i])
# print()
# assert len(active_set & difference_set) == 0
return data, labels, sorted(difference_set - active_set)
from inspect import signature import numpy as np from smtlearn.examples import ice_cream_problem from pywmi.plot import plot_data, plot_formula from pywmi.sample import uniform from pywmi.smt_check import evaluate import random from smtlearn.violations.core import RandomViolationsStrategy from smtlearn.k_cnf_smt_learner import KCnfSmtLearner from pywmi.smt_print import pretty_print random.seed(666) np.random.seed(666) domain, formula, name = ice_cream_problem() # plot_formula(None, domain, formula) data = uniform(domain, 100) labels = evaluate(domain, formula, data) learner = KCnfSmtLearner(3, 3, RandomViolationsStrategy(10)) initial_indices = random.sample(range(data.shape[0]), 20) learned_theory = learner.learn(domain, data, labels, initial_indices) print(pretty_print(learned_theory))
mutated_failing = []
mutated_failing.append(trainer_test_suite[0])
oracle_committee = []
timeout = time.time() + 60 * timeout
if debug: print("[INFO] Start learning (iteration: "+str(iteration)+")...")
while time.time() < timeout and n_human_labeled < max_labels:
if np.random.randint(0,100) < 50:
seed_input = mutated_failing[np.random.randint(len(mutated_failing))][:-1]
else:
seed_input = trainer_test_suite[0][:-1]
fuzzed_test_case = getMutatedTestCase(trainer_test_suite, seed_input)
predict_label = evaluate(domain,learned_model,np.array(fuzzed_test_case))
n_generated += 1
if False == ask_human(fuzzed_test_case[:-1], bug_dir, bug_prog, gold_prog):
n_failing += 1
all_fail_prob = (trainer_labels.size - np.sum(trainer_labels)) / trainer_labels.size
# If the current test_input is predicted as failing
# OR if we have mostly labled failing test cases and the current test input is predicted as passing
if predict_label == False or (trainer_labels.size > 10 and all_fail_prob > 0.9 and predict_label == True):
# ask human to label, and
human_label = ask_human(fuzzed_test_case[:-1], bug_dir, bug_prog, gold_prog)
# add to trainer_test_suite
trainer_test_suite = np.append(trainer_test_suite, [fuzzed_test_case], axis=0)
trainer_labels = np.append(trainer_labels, [human_label], axis=0)
# re-train automated oracle