Haoteng comments

Results 10 comments of


                                            Haoteng

Matlab called python Theano Module Wrong

Hi, Aviv! It still doesn't work on Matlab R2016b. I followed the instruction of Matlab documents, added the path into python. >> P = py.sys.path; >> P P = Python...

Data visualization

In the released code, we did not provide special visulization function or output for TensorBoard. You might have to save those predictions and plot them manually.

The hyperparameter setting under BJER4 scenario

For dataset like BJER4, it has relatively smaller graph size. The key issue here you may consider is the threshold for generating weighted matrix, and how to prevent over-fitting. Also,...

Whether SparseTensor supports element-wise operations (add, mul)?

Thanks for the quick response. I will certainly look into it.

How to convert input feature vector from 1 to many (ex. 7 or 9)

Please refer to the following issues, those discussions might be helpful for your question. https://github.com/VeritasYin/STGCN_IJCAI-18/issues/28 https://github.com/VeritasYin/STGCN_IJCAI-18/issues/29

how to make train dataset

You might refer this tutorial [[link]](https://keras.io/examples/timeseries/timeseries_traffic_forecasting/) to generate customized data.

Asking for the sensor location for PeMSD7-M

The station list is now available under the dataset folder.

batch training on linkx

> I am able to run the script without issues with `ogb==1.3.1`, `torch_geometric==1.7.2`, and `torch==1.8.1`. Can you check what your `ogb`, `torch_geometric`, and `torch` versions are and try switching it...

Stucked at Compiling `surel_gacc` on Macbook M1 chip

Hello, thank you for your interest in our work. Note that SUREL has not been tested on MacOS. 1) environment.yml is exported from Ubuntu 20.04, some packages may not work...

ogbn-arxiv text features don't match

Check Issue #222 might be helpful.