def setUp(self) -> None:
"""Set up deterioration tests."""
self.generator = ensure_torch_random_state(42)
self.reference = Nations()
self.num_training_triples = self.reference.training.num_triples
self.num_triples = (self.reference.training.num_triples +
self.reference.testing.num_triples +
self.reference.validation.num_triples)
def _pre_instantiation_hook(
self, kwargs: MutableMapping[str,
Any]) -> MutableMapping[str, Any]:
# create triples factory with inverse relations
kwargs = super()._pre_instantiation_hook(kwargs=kwargs)
kwargs["triples_factory"] = self.factory = Nations(
create_inverse_triples=True).training
return kwargs
def setUp(self) -> None:
"""Set up the test case with a triples factory and model."""
_, self.generator, _ = set_random_seed(42)
dataset = Nations(create_inverse_triples=self.create_inverse_triples)
self.factory = dataset.training
self.model = self.model_cls(self.factory,
embedding_dim=self.embedding_dim,
**(self.model_kwargs or {})).to_device_()
def test_custom_dataset_instance(self):
"""Test passing a pre-instantiated dataset to HPO."""
hpo_pipeline_result = self._help_test_hpo(
study_name="HPO with custom dataset instance",
dataset=Nations(
), # mock a "custom" dataset by using one already available
)
# Since custom data was passed, we can't store any of this
self.assertNotIn("dataset", hpo_pipeline_result.study.user_attrs)
self.assertNotIn("training", hpo_pipeline_result.study.user_attrs)
self.assertNotIn("testing", hpo_pipeline_result.study.user_attrs)
self.assertNotIn("validation", hpo_pipeline_result.study.user_attrs)
class TestDeterioration(unittest.TestCase):
"""Tests for deterioration workflow."""
def setUp(self) -> None:
"""Set up deterioration tests."""
self.generator = ensure_torch_random_state(42)
self.reference = Nations()
self.num_training_triples = self.reference.training.num_triples
self.num_triples = (self.reference.training.num_triples +
self.reference.testing.num_triples +
self.reference.validation.num_triples)
def test_deteriorate(self):
"""Test deterioration on integer values for ``n``."""
for n in [1, 2, 5, 10, 50, 100, 500, 1000]:
with self.subTest(n=n):
derived = self.reference.deteriorate(
n=n, random_state=self.generator)
self._help_check(derived)
self.assertEqual(
n, splits_steps(self.reference._tup(), derived._tup()))
self.assertEqual(1 - n / self.num_triples,
self.reference.similarity(derived),
msg="similarity")
self.assertEqual(1 - n / self.num_triples,
derived.similarity(self.reference),
msg="similarity")
def test_deteriorate_frac(self):
"""Test deterioration on fractional values for ``n``."""
for n_frac in [
1 / self.num_training_triples,
2 / self.num_training_triples,
5 / self.num_training_triples,
0.1,
0.2,
0.3,
]:
n = int(n_frac * self.num_training_triples)
with self.subTest(n=n, n_frac=n_frac):
derived = self.reference.deteriorate(
n=n, random_state=self.generator)
self._help_check(derived)
self.assertEqual(
n,
splits_steps(self.reference._tup(), derived._tup()),
msg="steps",
)
self.assertEqual(1 - n / self.num_triples,
self.reference.similarity(derived),
msg="similarity")
self.assertEqual(1 - n / self.num_triples,
derived.similarity(self.reference),
msg="similarity")
def _help_check(self, derived: Dataset):
self.assertIsNotNone(derived.validation)
self.assertEqual(self.num_training_triples,
self.reference.training.num_triples)
self.assertEqual(
self.num_triples,
sum((
derived.training.num_triples,
derived.testing.num_triples,
derived.validation.num_triples,
)),
msg="different number of total triples",
)
self.assertLess(derived.training.num_triples,
self.reference.training.num_triples)
