- Create a new environment with the following dependencies:
- Pytorch :
pip install torch torchvision torchaudio
- Pytorch geometric, according to your Pytorch version and CUDA version:
pip install pyg-lib torch-scatter torch-sparse torch-cluster torch-spline-conv torch-geometric -f https://data.pyg.org/whl/torch-{pytorch_version}+{cuda_version}.html
- Seaborn:
pip install seaborn
- Matplotlib:
pip install matplotlib
- Pandas:
pip install pandas
- torch-conformal package, from the root of the repository, run:
python setup.py install
- Pytorch :
- Run a Jupyter kernel on the environment
- Run the notebooks to reproduce the experiments
soroushzargar/DAPS
Implementations of methods proposed in the paper "Conformal Prediction Sets for Graph Neural Networks"
Jupyter Notebook