/examples/pytorch/stgcn_wave

Spatio-Temporal Graph Convolutional Networks

• Paper link: arXiv
• Author's code repo: https://github.com/VeritasYin/STGCN_IJCAI-18. Dependencies

---

• PyTorch 1.1.0+
• sklearn
• dgl
• tables

How to run

please get METR_LA dataset from this Google drive. and this Github repo

An experiment in default settings can be run with

python main.py

An experiment on the METR_LA dataset in customized settings can be run with

python main.py --lr --seed --disable-cuda --batch_size <batch-size> --epochs <number-of-epochs>

If one wishes to adjust the model structure, you can change the arguments control_str and channels

python main.py --control_str <control-string> --channels <n-input-channel> <n-hidden-channels-1> <n-hidden-channels-2> ... <n-output-channels>

<control-string> is a string of the following characters representing a sequence of neural network modules:

• T: representing a dilated temporal convolution layer, working on the temporal dimension. The dilation factor is always twice as much as the previous temporal convolution layer.
• S: representing a graph convolution layer, working on the spatial dimension. The input channels and output channels are the same.
• N: a Layer Normalization.

The argument list following --channels represents the output channels on each temporal convolution layer. The list should have N + 1 elements, where N is the number of T's in <control-string>.

The activation function between two layers are always ReLU.

For example, the following command

python main.py --control_str TNTSTNTST --channels 1 16 32 32 64 128

specifies the following architecture:

+------------------------------------------------------------+
|                          Input                             |
+------------------------------------------------------------+
|  1D Conv, in_channel = 1, out_channel = 16, dilation = 1   |
+------------------------------------------------------------+
|                   Layer Normalization                      |
+------------------------------------------------------------+
|  1D Conv, in_channel = 16, out_channel = 32, dilation = 2  |
+------------------------------------------------------------+
|       Graph Conv, in_channel = 32, out_channel = 32        |
+------------------------------------------------------------+
|  1D Conv, in_channel = 32, out_channel = 32, dilation = 4  |
+------------------------------------------------------------+
|                   Layer Normalization                      |
+------------------------------------------------------------+
|  1D Conv, in_channel = 32, out_channel = 64, dilation = 8  |
+------------------------------------------------------------+
|       Graph Conv, in_channel = 64, out_channel = 64        |
+------------------------------------------------------------+
| 1D Conv, in_channel = 64, out_channel = 128, dilation = 16 |
+------------------------------------------------------------+

Results

python main.py

METR_LA MAE: ~5.76