RLProject

This application is used to simulate an Agent-Environment Interaction.

Code Structure:

  • CartPole : contains CartPole Domain
    • CartPole.py : defines the state and dynamics of cartpole domain
    • CPEpisode.py : runs an episode of cartpole
    • runCartpole.py : python file that needs to be executed
  • Gridworld : contains Gridworld Domain
    • Gridworld.py : contains State Definition including policy, transition, rewards and getNextState.
    • ValueIteration.py : contains the value iteration algorithm to find optimal policy.
    • GWEpisode.py : runs one episode.
    • CommonFunctions.py : contains a few necessary common functions.
    • runGridworld.py : python file that needs to be executed

Setup:

  • Requires python3 to be installed.
  • Python version used to test is Python 3.11.5
  • Command: python runCartpole.py inside CartPole Directory
  • Command: python runGridworld.py inside Gridworld Directory