commonpower.control.wrappers

Wrappers to adjust API in environments.py to different RL training algorithms.

Functions

ctrl_dict_to_list

Transforms the orginal dict of the controller assignments to a list of lists.

list_to_ctrl_dict

Reverses the transform_to_ordered_dict_keys function.

recursive_items

Recursive extraction of all values in a nested dictionary or gym.spaces.Dict

Classes

DeploymentWrapper

Wrapper to standardize the deployment within CommonPower.

MultiAgentWrapper

Wrapper to standardize ControlEnv to the API for MAPPO/IPPO implementation of the on-policy repository (https://github.com/marlbenchmark/on-policy/tree/main/onpolicy).

RecordTransitionsWrapper

Wrapper for recording transition tuples (s,a,s',r) either to current disk or to a data base.

SingleAgentWrapper

Wrapper to standardize ControlEnv to the API for single-agent RL training with any RL algorithm from the StableBaselines 3 repository.

WrapperStack