Skip to content

A programming framework for agentic AI

Use Ray _BackwardsCompatibleNumpyRng if possible (#421)

Antoni Baum113539545c0c
diff --git a/flaml/tune/sample.py b/flaml/tune/sample.py
index e03edf2ef..113006fe1 100644
--- a/flaml/tune/sample.py
+++ b/flaml/tune/sample.py
@@ -35,46 +35,48 @@ except AttributeError:
logger = logging.getLogger(__name__)
+try:
+ from ray.tune.sample import _BackwardsCompatibleNumpyRng
+except ImportError:
+ class _BackwardsCompatibleNumpyRng:
+ """Thin wrapper to ensure backwards compatibility between
+ new and old numpy randomness generators.
+ """
-class _BackwardsCompatibleNumpyRng:
- """Thin wrapper to ensure backwards compatibility between
- new and old numpy randomness generators.
- """
-
- _rng = None
+ _rng = None
- def __init__(
- self,
- generator_or_seed: Optional[
- Union["np_random_generator", np.random.RandomState, int]
- ] = None,
- ):
- if generator_or_seed is None or isinstance(
- generator_or_seed, (np.random.RandomState, np_random_generator)
+ def __init__(
+ self,
+ generator_or_seed: Optional[
+ Union["np_random_generator", np.random.RandomState, int]
+ ] = None,
):
- self._rng = generator_or_seed
- elif LEGACY_RNG:
- self._rng = np.random.RandomState(generator_or_seed)
- else:
- self._rng = np.random.default_rng(generator_or_seed)
-
- @property
- def legacy_rng(self) -> bool:
- return not isinstance(self._rng, np_random_generator)
-
- @property
- def rng(self):
- # don't set self._rng to np.random to avoid picking issues
- return self._rng if self._rng is not None else np.random
-
- def __getattr__(self, name: str) -> Any:
- # https://numpy.org/doc/stable/reference/random/new-or-different.html
- if self.legacy_rng:
- if name == "integers":
- name = "randint"
- elif name == "random":
- name = "rand"
- return getattr(self.rng, name)
+ if generator_or_seed is None or isinstance(
+ generator_or_seed, (np.random.RandomState, np_random_generator)
+ ):
+ self._rng = generator_or_seed
+ elif LEGACY_RNG:
+ self._rng = np.random.RandomState(generator_or_seed)
+ else:
+ self._rng = np.random.default_rng(generator_or_seed)
+
+ @property
+ def legacy_rng(self) -> bool:
+ return not isinstance(self._rng, np_random_generator)
+
+ @property
+ def rng(self):
+ # don't set self._rng to np.random to avoid picking issues
+ return self._rng if self._rng is not None else np.random
+
+ def __getattr__(self, name: str) -> Any:
+ # https://numpy.org/doc/stable/reference/random/new-or-different.html
+ if self.legacy_rng:
+ if name == "integers":
+ name = "randint"
+ elif name == "random":
+ name = "rand"
+ return getattr(self.rng, name)
RandomState = Union[