Skip to content

Ceil integer attributes in undiscretize when clip_to_range is False #233

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
copybara-service merged 1 commit into from
Oct 6, 2026
Merged

Ceil integer attributes in undiscretize when clip_to_range is False #233

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions dpsynth/local_mode/vectorized_transformations.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -218,8 +218,10 @@ def undiscretize(
else:
raise ValueError(f'Unsupported interval_handling: {handling}')

if attribute_domain.dtype == 'int' and attribute_domain.clip_to_range:
result = np.ceil(result).astype(int)
if attribute_domain.dtype == 'int':
result = np.ceil(result)
if attribute_domain.clip_to_range or not np.isnan(sentinel):
result = result.astype(int)
return result


Expand Down
21 changes: 21 additions & 0 deletions tests/local_mode/vectorized_transformations_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -298,6 +298,27 @@ def test_custom_sentinel_sample(self):
)
self.assertEqual(result[0], -1)

@parameterized.parameters(('midpoint', None), ('sample', -1))
def test_integer_dtype_no_clip(self, interval_handling, sentinel):
rng = np.random.default_rng(0)
attr = domain.NumericalAttribute(
min_value=0,
max_value=3,
dtype='int',
clip_to_range=False,
sentinel=sentinel,
interval_handling=interval_handling,
)
result = vectorized_transformations.undiscretize(
np.array([0, 1, 2, 3, 4]), np.array([0.0, 1.0, 2.0]), attr, rng=rng
)
np.testing.assert_array_equal(result[1:], [0, 1, 2, 3])
if sentinel is None:
self.assertTrue(np.isnan(result[0]))
else:
self.assertTrue(np.issubdtype(result.dtype, np.integer))
self.assertEqual(result[0], sentinel)


class MergeRareValuesTest(absltest.TestCase):

Expand Down
Loading