In this folder, 'code_dataset.py' file achieves pipeline of constructing AST-based program graphs.
The dataset of experiment is stored in this folder.
In this folder, 'explaining_subgraphx.py' file achieves the entire pipeline of graph construction, GNN-based model training and GNN interpretation with SubgraphX. Go to this explanation file and run the following script.
python explaining_subgraphx.py
In this folder, all the visualized interpretation results (png files) are stored.
The detailed versions of dependencies are recorded in requirement.txt file, run the following script to start.
pip install -r requirements.txt