Skip to content

Improving MD #34

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
BaseMax wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Improving MD #34

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 16 additions & 6 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,27 +1,33 @@
# MTGNN

This is a PyTorch implementation of the paper: [Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks](https://arxiv.org/abs/2005.11650), published in KDD-2020.

## Requirements

The model is implemented using Python3 with dependencies specified in requirements.txt

## Data Preparation

### Multivariate time series datasets

Download Solar-Energy, Traffic, Electricity, Exchange-rate datasets from [https://github.com/laiguokun/multivariate-time-series-data](https://github.com/laiguokun/multivariate-time-series-data). Uncompress them and move them to the data folder.

### Traffic datasets

Download the METR-LA and PEMS-BAY dataset from [Google Drive](https://drive.google.com/open?id=10FOTa6HXPqX8Pf5WRoRwcFnW9BrNZEIX) or [Baidu Yun](https://pan.baidu.com/s/14Yy9isAIZYdU__OYEQGa_g) provided by [Li et al.](https://github.com/liyaguang/DCRNN.git) . Move them into the data folder.

```

# Create data directories

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

# METR-LA

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

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

python generate_training_data.py --output_dir=data/PEMS-BAY --traffic_df_filename=data/pems-bay.h5
```

## Model Training
Expand All @@ -32,22 +38,23 @@ python generate_training_data.py --output_dir=data/PEMS-BAY --traffic_df_filenam

```
python train_single_step.py --save ./model-solar-3.pt --data ./data/solar_AL.txt --num_nodes 137 --batch_size 4 --epochs 30 --horizon 3
#sampling
# sampling
python train_single_step.py --num_split 3 --save ./model-solar-sampling-3.pt --data ./data/solar_AL.txt --num_nodes 137 --batch_size 16 --epochs 30 --horizon 3
```
* Traffic

* Traffic

```
python train_single_step.py --save ./model-traffic3.pt --data ./data/traffic.txt --num_nodes 862 --batch_size 16 --epochs 30 --horizon 3
#sampling
# sampling
python train_single_step.py --num_split 3 --save ./model-traffic-sampling-3.pt --data ./data/traffic --num_nodes 321 --batch_size 16 --epochs 30 --horizon 3
```

* Electricity

```
python train_single_step.py --save ./model-electricity-3.pt --data ./data/electricity.txt --num_nodes 321 --batch_size 4 --epochs 30 --horizon 3
#sampling
# sampling
python train_single_step.py --num_split 3 --save ./model-electricity-sampling-3.pt --data ./data/electricity.txt --num_nodes 321 --batch_size 16 --epochs 30 --horizon 3
```

Expand All @@ -58,12 +65,15 @@ python train_single_step.py --save ./model/model-exchange-3.pt --data ./data/exc
#sampling
python train_single_step.py --num_split 3 --save ./model-exchange-3.pt --data ./data/exchange_rate.txt --num_nodes 8 --subgraph_size 2 --batch_size 16 --epochs 30 --horizon 3
```

### Multi-step

* METR-LA

```
python train_multi_step.py --adj_data ./data/sensor_graph/adj_mx.pkl --data ./data/METR-LA --num_nodes 207
```

* PEMS-BAY

```
Expand Down