Tensor utilities, reinforcement learning, and more! Designed to make research easier with low-level abstractions for common operations.
For the top-level utilities, import like so:
import pyoneer as pynr
pynr.math.rescale(...)For the larger sub-modules, such as reinforcement learning, we recommend:
import pyoneer.rl as pyrl
loss_fn = pyrl.losses.PolicyGradient(...)In general, the Pyoneer API tries to adhere to the TensorFlow 2.0 API.
pynr.activations)pynr.activations.swishpynr.debugging)pynr.debugging.Stopwatchpynr.distributions)pynr.distributions.MultiCategoricalpynr.initializers)pynr.initializers.SoftplusInversepynr.layers)pynr.layers.Swishpynr.layers.OneHotEncoderpynr.layers.AngleEncoderpynr.layers.Nestpynr.layers.ConcreteDropoutpynr.manip)pynr.manip.flattenpynr.manip.batched_indexpynr.manip.pad_or_truncatepynr.manip.shiftpynr.math)pynr.math.to_radianspynr.math.to_degreespynr.math.to_cartesianpynr.math.to_polarpynr.math.RADIANS_TO_DEGREESpynr.math.DEGREES_TO_RADIANSpynr.math.isclosepynr.math.safe_dividepynr.math.rescalepynr.math.normalizepynr.math.denormalizepynr.math.sigmoid_inversepynr.metrics)pynr.metrics.mapepynr.metrics.smapepynr.metrics.MAPEpynr.metrics.SMAPEpynr.moments)pynr.moments.range_momentspynr.moments.StaticMomentspynr.moments.StreamingMomentspynr.moments.ExponentialMovingMomentspynr.schedules)pynr.schedules.CyclicSchedulepynr.regularizers)pynr.regularizers.DropoutL2pynr.rl)Utilities for reinforcement learning.
pynr.rl.losses)pynr.rl.losses.policy_gradientpynr.rl.losses.policy_entropypynr.rl.losses.clipped_policy_gradientpynr.rl.losses.PolicyGradientpynr.rl.losses.PolicyEntropypynr.rl.losses.ClippedPolicyGradientpynr.rl.targets)pynr.rl.targets.DiscountedReturnspynr.rl.targets.GeneralizedAdvantagespynr.rl.strategies)pynr.rl.strategies.EpsilonGreedypynr.rl.strategies.Modepynr.rl.strategies.Samplepynr.rl.wrappers)pynr.rl.wrappers.ObservationCoordinatespynr.rl.wrappers.ObservationNormalizationpynr.rl.wrappers.Batchpynr.rl.wrappers.ProcessThere are a few options for installation:
(Recommended) Install with pipenv:
pipenv install fomoro-pyoneerInstall locally for development with pipenv:
git clone https://github.com/fomorians/pyoneer.git
cd pyoneer
pipenv install
pipenv shellThere are a few options for testing:
Run all tests:
python -m unittest discover -bfp '*_test.py'Run specific tests:
python -m pyoneer.math.logical_ops_testFile an issue following the ISSUE_TEMPLATE. If the issue discussion warrants implementation, then submit a pull request from a branch describing the feature. This will eventually get merged into master after a few rounds of code review.