Skip to content
This repository has been archived by the owner on Jul 29, 2024. It is now read-only.

Latest commit

 

History

History
46 lines (30 loc) · 1.79 KB

README.md

File metadata and controls

46 lines (30 loc) · 1.79 KB
Raw

RLLIB+SUMO Utils

Python3 library able to connect the RLLIB framework with the SUMO simulator.

Contact: Lara CODECA [[email protected]]

This program and the accompanying materials are made available under the terms of the Eclipse Public License 2.0 which is available at http://www.eclipse.org/legal/epl-2.0.

Minimum Requirements

Tested with

  • Eclipse SUMO Version 1.9.2
    • With 1.9.2 the options
      • scenario_config['sumo_config']['sumo_connector'] = 'libsumo' with scenario_config['sumo_config']['trace_file'] = True does not work, but the bugh has already been fixed in SUMO issue 8671. If you need to enable SUMO tracing with 1.9.2, you need to use scenario_config['sumo_config']['sumo_connector'] = 'traci'
  • RLlib Version 1.3.0

Installation

  • Install: pip3 install . from the root directory, or python3 setup.py install
  • Development install: pip3 install -e . or python3 setup.py develop

Note: rtree library

To use rtree, libspatialindex-dev is required to be installed.

[In Ubuntu/Debian] sudo apt install libspatialindex-dev python-rtree and then pip install rtree

Example

  • Given the under development status of the project, some examples are provided.
    • example/scenario
      • Simple SUMO scenario.
    • example/marlenvironment.py
      • Example of MARL environment implemented using RLLIB (SUMOTestMultiAgentEnv)
    • example/train.py
      • Example of PPO trainer using SUMOTestMultiAgentEnv
      • How to: python3 train.py

Docker Environment

See RLLIB SUMO Docker for details on my development and learning environment.