OR-Gym provides a suite of OpenAI Gym-compatible environments for classic operations research (OR) problems, targeting researchers and practitioners in both OR and reinforcement learning (RL). It aims to bridge the gap between these fields by offering accessible simulations for developing and benchmarking RL agents against traditional OR techniques.

How It Works

The library implements various OR problems as discrete or continuous state/action spaces within the Gym API. This allows RL algorithms to interact with these problems as if they were standard simulation environments. The design facilitates direct comparison of RL agent performance against established OR heuristics and solvers on well-defined problem instances.

Quick Start & Requirements

• Install via pip: $ pip install or-gym
• Or from GitHub: git clone https://github.com/hubbs5/or-gym.git && cd or-gym && pip install -e .
• Requires Python 3.5+.
• Example notebook inv-management-quickstart.ipynb demonstrates usage.
• Ray 1.0.0 is required for RL algorithm examples.

Highlighted Details

• Environments cover Knapsack (unbounded, 0-1, bounded, stochastic), Bin Packing, Newsvendor, Virtual Machine Packing, Vehicle Routing, Inventory Management, Network Management, and Traveling Salesman Problems.
• Includes examples for action masking with RLlib and improving supply chains with DRL.
• Provides a citation for academic use.

Maintenance & Community

• The project was published in 2020. No recent activity or community links (Discord/Slack) are mentioned in the README.

Licensing & Compatibility

• The README does not explicitly state a license. Compatibility for commercial use or closed-source linking is undetermined.

Limitations & Caveats

The README does not specify a license, which may impact commercial adoption. Community support and recent maintenance status are not detailed, raising potential concerns about the project's long-term viability.