Discrete Graph Structure Learning for Forecasting Multiple Time Series

This is a PyTorch implementation of the paper "Discrete Graph Structure Learning for Forecasting Multiple Time Series", ICLR 2021.

Installation

Install the dependency using the following command:

pip install -r requirements.txt

• torch
• scipy>=0.19.0
• numpy>=1.12.1
• pandas>=0.19.2
• pyyaml
• statsmodels
• tensorflow>=1.3.0
• tables
• future

Data Preparation

The traffic data files for Los Angeles (METR-LA) and the Bay Area (PEMS-BAY) are put into the data/ folder. They are provided by DCRNN.

Run the following commands to generate train/test/val dataset at data/{METR-LA,PEMS-BAY}/{train,val,test}.npz.

# Unzip the datasets
unzip data/metr-la.h5.zip -d data/
unzip data/pems-bay.h5.zip -d data/

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

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

# PEMS-BAY
python -m scripts.generate_training_data --output_dir=data/PEMS-BAY --traffic_df_filename=data/pems-bay.h5

Train Model

When you train the model, you can run:

# Use METR-LA dataset
python train.py --config_filename=data/model/para_la.yaml --temperature=0.5

# Use PEMS-BAY dataset
python train.py --config_filename=data/model/para_bay.yaml --temperature=0.5

Hyperparameters can be modified in the para_la.yaml and para_bay.yaml files.

Design your own model

You can directly modify the model in the "model/pytorch/model.py" file.

Citation

If you use this repository, e.g., the code and the datasets, in your research, please cite the following paper:

@article{shang2021discrete,
  title={Discrete Graph Structure Learning for Forecasting Multiple Time Series},
  author={Shang, Chao and Chen, Jie and Bi, Jinbo},
  journal={arXiv preprint arXiv:2101.06861},
  year={2021}
}

Acknowledgments

DCRNN-PyTorch, GCN, NRI and LDS-GNN.