Quantum Cartpole

Introduction

Python implementation of a quantum cartpole environment which can be used for benchmarking and developing of agents as outline in the paper. The implementation adheres to the OpenAI Gym standard.

Under the MIT lisence.

Installation

To get started with the Quantum Cartpole environment, you can install it by simply cloning this repository:

git clone https://github.com/condensedAI/quantum-cartpole.git

Basic Usage

The Quantum Cartpole environment can be used as follows. It offers a choice of three different potentials: quadratic, cosine, and quartic, and can be switched between classical and quantum environment.

import gym
import gym_qcart
env = gym.make('qcart-v0', potential = 'quadratic', system = 'quantum')

To run examples for training and testing the agents, you can use the following commands:

For training with a quadratic potential:

python -m example.train

For testing:

python -m example.test quadratic

Reference

When you cite this repository, please use the following:

@software{meinerz_2023_10060570,
  author       = {Meinerz, Kai},
  title        = {Quantum cartpole environment},
  month        = nov,
  year         = 2023,
  publisher    = {Zenodo},
  version      = {v0.1.0},
  doi          = {10.5281/zenodo.10060570},
  url          = {https://doi.org/10.5281/zenodo.10060570}
}