Elevator dispatch environment for Gym

This is a project for Reinforcement Learning Course in UniNE

Directory Structure

• Elevators
  • envs
    • simple_elevator.py: contains the core logic of the environment
• Solver
  • BaseSolver.py: Base class for all the solvers
  • FIFO.py: implement FIFO to solve elevator dispatch
  • LOOK.py: implement LOOK to solve elevator dispatch
• visualization: you can use python *.py -h for all the python files in this folder to check how to use it.
  • event_plot.py: parse the render output from this environment and plot a beautiful event timeline
  • load_plot.py: parse the render output from this environment and plot a beautiful event timeline, the darker the color is, the larger load elevator has.
  • rewards_boxplot.py: You give a list of rewards to the program, it will produce a box plot to show the mean and variance of the rewards.
  • filling_plot.py: You give losses to it, it will produce a nice figure with running means and variance filling area.
  • ppo_handle.py: The PPO rewards are accumulated, I use this to recover the original rewards
• RL_elevator.ipynb: contains the core logic of training RL models

Installation

To install your new environment, run the following commands:

pip install -r requirements.txt
cd Elevators
pip install -e .

Contributing

If you would like to contribute, follow these steps:

• Fork this repository
• Clone your fork
• Set up pre-commit via pre-commit install

PRs may require accompanying PRs in the documentation repo.