If this is not a strict requirement, we should perhaps remove this or make it more lenient?
================================== FAILURES ===================================
______________ test_trafo_1[search_space2-memory_warm_start2-4] _______________
random_state = 4
search_space = {'min_samples_split': [np.int64(22), np.int64(23), np.int64(24), np.int64(25), np.int64(26), np.int64(27), ...]}
memory_warm_start = min_samples_split score
0 22 22
1 23 23
2 24 24
3 ... 27
6 28 28
7 29 29
8 30 30
9 31 31
@pytest.mark.parametrize("random_state", random_state_para_0)
@pytest.mark.parametrize("search_space, memory_warm_start", search_space_para_1)
def test_trafo_1(random_state, search_space, memory_warm_start):
search_space = search_space
memory_warm_start = memory_warm_start
c_time_0 = time.perf_counter()
hyper = Hyperactive()
hyper.add_search(
objective_function_1,
search_space,
n_iter=10,
random_state=random_state,
initialize={"random": 1},
)
hyper.run()
d_time_0 = time.perf_counter() - c_time_0
c_time_1 = time.perf_counter()
hyper = Hyperactive()
hyper.add_search(
objective_function_1,
search_space,
n_iter=10,
random_state=random_state,
initialize={"random": 1},
memory_warm_start=memory_warm_start,
)
hyper.run()
d_time_1 = time.perf_counter() - c_time_1
> assert d_time_1 < d_time_0 * 0.5
E assert 1.027204499999982 < (1.9211578000000031 * 0.5)
test_trafo_1fails sporadically with a timings check. Example traceback below.If this is not a strict requirement, we should perhaps remove this or make it more lenient?