/MetroGNN

Primary LanguagePythonMIT LicenseMIT

MetroGNN

Framework

Loading Model Overview

Episode

image-20231007184646501

Installation

Environment

  • **Tested OS: **Linux
  • Python >= 3.8
  • PyTorch == 1.10.1
  • Tensorboard

Dependencies:

  1. Install PyTorch 1.10.1 with the correct CUDA version.
  2. Set the following environment variable to avoid problems with multiprocess trajectory sampling: 
    export OMP_NUM_THREADS=1

Training

You can train your own models using the provided config in metro/cfg:

python -m metro.train --city_name city_name --cfg demo

You can replace demo to train other cfgs.

The results are saved in result/city_name/cfg/seed

Convergence

Our models MetroGNN, usually converges with less than 100 iterations, which takes about 14 hours on a single server with an Nvidia GeForce 2080Ti GPU.

Result

image-20231007184747711

image-20231007184843088

  • We compared MetroGNN to other baselines at different budgets, with the metrics being OD flow satisfaction.
  • If two regions can be connected by a metro network, the product of the OD flow and the transportation efficiency (the ratio of straight-line distance to path distance) between the two regions is considered to be the OD flow satisfaction.

License

Please see the license for further details.

Note

The implemention is based on Transform2Act.