def test_pseudorandom_split(all_values):
split_list = [0.3, 0.4, 0.1, 0.0, 0.2]
values_num = len(all_values)
for idx in range(len(split_list)):
test_predicate = in_pseudorandom_split(split_list, idx, 'volume_guid')
included_values = set()
for val in all_values:
if test_predicate.do_include({'volume_guid': val}):
included_values.add(val)
expected_num = values_num * split_list[idx]
assert pytest.approx(len(included_values),
expected_num * 0.1) == expected_num
def test_pseudorandom_split_on_integer_field():
split_list = [0.3, 0.4, 0.1, 0.0, 0.2]
int_values = list(range(1000))
values_num = len(int_values)
for idx, _ in enumerate(split_list):
test_predicate = in_pseudorandom_split(split_list, idx,
'int_partitioning_field')
included_values = set()
for val in int_values:
if test_predicate.do_include({'int_partitioning_field': val}):
included_values.add(val)
expected_num = values_num * split_list[idx]
assert pytest.approx(len(included_values),
expected_num * 0.1) == expected_num
def test_pseudorandom_split(self):
split_list = [0.3, 0.4, 0.1, 0.0, 0.2]
values_num = len(PredicatesTest.all_values)
for idx in range(len(split_list)):
test_predicate = in_pseudorandom_split(split_list, idx,
'volume_guid')
included_values = set()
for val in PredicatesTest.all_values:
if test_predicate.do_include({'volume_guid': val}):
included_values.add(val)
expected_num = values_num * split_list[idx]
self.assertAlmostEqual(len(included_values),
expected_num,
delta=expected_num * 0.1)