Structure:

• data: including PEMSD4 and PEMSD8 dataset used in our experiments, which are released by and available at ASTGCN.

• lib: contains self-defined modules for our work, such as data loading, data pre-process, normalization, and evaluate metrics.

• model: implementation of our model

The code for training and loading data are forked from AGCRN.

Requirements

python=3.6
pytorch=1.9.0
numpy=1.19
argparse
configparser

Training

To train the model, you can simply run the following code in terminal:

cd model
python Run.py

We provide pretrained model, you can run the following code to test:

python Run.py --mode test