This code contains a PyTorch implementation of "Improving Expressive Power of Spectral Graph Neural Networks with Eigenvalue Correction" (AAAI 2024)

• pytorch 1.8.1
• numpy 1.18.1
• torch-geometric 1.6.3
• tqdm 4.59.0
• scipy 1.6.2
• seaborn 0.11.1
• scikit-learn 0.24.1

You can run the following Command:

python histogram.py

You can run the following Command:

python distinct_eigvalues.py

Run the following command to perform eigendecomposition on all ten datasets, and will print the time required for each eigendecomposition.

python eigendecomposition.py

The eigenvalues and eigenvectors are stored in the data directory.

You can run the following script directly:

sh EC-Bern.sh

or run the following Command

sh EC-GPR.sh