fractal13 / uniform-coins-environment

Demo of a simple searchable gymnasium environment.

Uniform Coins

This is a simple environment to demonstrate creation of a gymnasium environment that has a model to support classic search as well as being used for reinforcement learning (the primary purpose of most gymnasium environments).

Farama has a turtorial on creating your own custom environment. Those instructions were used, along with the requirements of having an environment model that can be plugged into classical tree search.

To be sure that all pip prerequisites are installed first, use the Makefile in prerequisites.

The gymnasium environment is contained completely in uniform-coins. Use the Makefile in that directory to install the module in your local pip. Note that this does not copy the module, it sets a link to this location. This is convenient for making updates after installing it.

demo-agents has a random agent and a iterative deepening search agent, to demonstrate the use of the environment and the model side-by-side in running the environment and searching for solutions.