rlai | This is a Python implementation of concepts and algorithms described in “Reinforcement Learning: An Introduction” (Sutton and Barto, 2018, 2nd edition).

Home > Environments

rlai.core.environments.mancala.Mancala

Environment for the mancala game. This is a simple game with many rule variations, and it provides a greater
    challenge in terms of implementation and state-space size than the gridworld. I have implemented a fairly common
    variation summarized below.

    * One row of 6 pockets per player, each starting with 4 seeds.
    * Landing in the store earns another turn.
    * Landing in own empty pocket steals.
    * Game terminates when a player's pockets are clear.
    * Winner determined by store count.

    A couple hours of Monte Carlo optimization explores more than 1 million states when playing against an equiprobable
    random opponent.

rlai.core.environments.mdp.ContinuousMdpEnvironment

MDP environment in which states and actions are continuous and multidimensional.

rlai.core.environments.network.TcpMdpEnvironment

An MDP environment served over a TCP connection from an external source (e.g., a simulation environment running as
    a separate program).

rlai.core.environments.openai_gym.Gym

Generalized Gym environment. Any OpenAI Gym environment can be executed by supplying the appropriate identifier.

rlai.core.environments.robocode.RobocodeEnvironment

Robocode environment. The Java implementation of Robocode runs alongside the current environment, and a specialized
    robot implementation on the Java side makes TCP calls to the present Python class to exchange action and state
    information.

rlai.core.environments.robocode_continuous_action.RobocodeEnvironment

Robocode environment. The Java implementation of Robocode runs alongside the current environment, and a specialized
    robot implementation on the Java side makes TCP calls to the present Python class to exchange action and state
    information.