# # Copyright (c) 2021 The Markovflow Contributors. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # """ Module containing type aliases, attributes and helper functions. .. autodata:: SampleShape """ import os from typing import List, Tuple, Union import tensorflow_probability as tfp [docs]def ordered(): """ A bijector to be used when parameterising inducing points so that we ensure they remain ordered. TensorFlow Probability unfortunately got their naming the wrong way around, see https://github.com/tensorflow/probability/issues/765 - we need to construct Invert(Ordered()) to obtain a Parameter that is always ordered! Note: when using TensorFlow Probability >= 0.12.0 this can be replaced with `tfp.bijectors.Ascending`. """ return tfp.bijectors.Invert(tfp.bijectors.Ordered()) [docs]SampleShape = Union[Tuple, List, int] """ A type that is either a ``Tuple``, a ``List``, or an ``int``. """ [docs]APPROX_INF = 1e10 """ Constant defining an appropriate level of approximate inference. """ [docs]AUTO_NAMESCOPE = "AUTO_NAMESCOPE" """ Name of environmental variable which if set enables a well structured TensorFlow graph for tensorboard debugging. In rare cases may cause issues, and will clutter the stacktrace so off by default. """ [docs]def auto_namescope_enabled() -> bool: """ Return `True` if autonamescoping is enabled. See the description of AUTO_NAMESCOPE.""" return True if os.environ.get(AUTO_NAMESCOPE) else False