This program is designed to perform node classification using a Graph Neural Network (GNN) model. It supports the R-GCN (Relational Graph Convolutional Network) architecture and provides options for training the model, performing graph explanation, and evaluating the performance metrics.

To run this program, please follow these steps:

1. Clone the repository:

git clone <repository-url>

2. Install the required dependencies:

pip install -r requirements.txt

3. Set up the dataset:

By default, the program uses the "mutag" dataset. If you want to use a different dataset, make sure to update the --dataset argument in the command-line or modify the default value in the code.

4. Run the program:

python main.py

The program accepts the following command-line arguments:

• --dropout: Dropout probability for regularization (default: 0).
• --n-hidden: Number of hidden units in the model (default: 16).
• --gpu: GPU device index to use. Set to -1 for CPU (default: -1).
• --lr: Learning rate for the optimizer (default: 0.01).
• --n-bases: Number of filter weight matrices. Use -1 to use all (default: -1).
• --n-layers: Number of propagation rounds (default: 2).
• -e, --n-epochs: Number of training epochs (default: 50).
• -d, --dataset: Dataset to use (default: "mutag").
• --model_path: Path to save the trained model (default: "data/saved_model.pt").
• --explain_graph: Set to True if graph explanation is needed (default: False).
• --print_metrics: Set to True to print evaluation metrics (default: False).
• --node_index: Provide node index if a single node index output is expected (default: None).
• --l2norm: L2 norm coefficient (default: 0).
• --use-self-loop: Include self-feature as a special relation (default: False).

To train the model and evaluate its performance:

python main.py

To enable graph explanation:

python main.py --explain_graph True

To print evaluation metrics:

python main.py --print_metrics True

To print for a single node index only

python main.py --node_index 1

This program is licensed under the MIT License.