ethicalgardeners.gardenersenvΒΆ

The GardenersEnv module provides the main simulation environment for the Ethical Gardeners reinforcement learning platform.

This module implements the PettingZoo AECEnv interface, serving as the primary entry point of the simulation. It coordinates all simulation components:

  1. World representation and state management (gridworld)

  2. Agent actions and interactions (ActionHandler)

  3. Observation generation (observation)

  4. Reward calculation (RewardFunctions)

  5. Metrics tracking (MetricsCollector)

  6. Visualization rendering (renderer)

The environment is highly configurable through Hydra configuration files.

members:

Classes

GardenersEnv(random_generator, grid_world, ...)

Main environment class implementing the PettingZoo AECEnv interface.