/gym_dataset_recorder

Package for recording Transitions in OpenAI Gym Environments.

Primary LanguagePythonMIT LicenseMIT

gym-dataset-recorder

Package for recording transitions in OpenAI Gym Environments and generating datasets from them.

Installation

git clone https://github.com/alxschwrz/gym-dataset-recorder.git
cd gym-dataset-recorder
pip3 install -e .

Run example

python3 example.py

Recorded data is stored as csv-file containing the following information:

  • observations: An [N x observation] dimensional array of observations.
  • actions: An [N x action] dimensional array of actions.
  • terminals: An [N x dimensional] array of episode termination flags. This is true when episodes end due to termination conditions such as falling over.
  • rewards: An [N] dimensional array of rewards.
  • next_observations: (optional) An [N x observation] dimensional array of the following observations.
  • goal: (optional) An [N] dimensional array of the current goals
  • infos: An [N] dimensional dict containing optional task-specific debugging information.