Traffic Transformer

A pytorch implementation of Traffic Transformer for traffic forecasting.Now the corresponding paper is available online at (https://ieeexplore.ieee.org/document/9520129)

Acknowledgement

Thank to the authors of Graph WaveNet and DCRNN. My work stands on their basic code and data.

Requirements

python 3
see requirements.txt

Data

Step1: Download METR-LA data from Google Drive or Baidu Yun links provided by DCRNN.

Step2: Process raw data

# Create data directories
mkdir -p data/{METR-LA}

# METR-LA
python generate_training_data.py --output_dir=data/METR-LA --traffic_df_filename=data/metr-la.h5

Train

python train.py

Test

python test.py

Citation

If you find this work helpful, please kindly cite our paper

@ARTICLE{9520129,
  author={Yan, Haoyang and Ma, Xiaolei and Pu, Ziyuan},
  journal={IEEE Transactions on Intelligent Transportation Systems}, 
  title={Learning Dynamic and Hierarchical Traffic Spatiotemporal Features With Transformer}, 
  year={2022},
  volume={23},
  number={11},
  pages={22386-22399},
  doi={10.1109/TITS.2021.3102983}}

R0oup1iao/Traffic-Transformer