Tutorials¶

This page lists the runnable scripts in the examples/ folder.

Run examples from the repository root:

make setup

# Optional setup
make setup-tensorboard

make run examples/hyperlink_prediction/nhp.py

Under examples/, the scripts are organized by task. The following table lists all the runnable scripts along with a brief description.

Example	What it demonstrates	Run
Dataset management
load_dataset_from_external_source.py	Load dataset from external source	`make run examples/dataset/load_dataset_from_external_source.py`
Node and hyperedge enrichment
hyperedge_enricher.py	Hyperedge enrichment: weights (degree) + hyperedge attributes	`make run examples/enrichment/hyperedge_enricher.py`
node_enricher.py	Node feature enrichment: Laplacian positional encoding (LPE) + Node2Vec	`make run examples/enrichment/node_enricher.py`
Hyperlink prediction models
common_neighbors.py	Common Neighbors HLP baseline on `AlgebraDataset` (negative sampling)	`make run examples/hyperlink_prediction/common_neighbors.py`
gcn.py	GCN HLP pipeline on `AlgebraDataset` (negative sampling + LPE enricher)	`make run examples/hyperlink_prediction/gcn.py`
hgnn.py	HGNN HLP pipeline on `AlgebraDataset` (negative sampling + LPE enricher)	`make run examples/hyperlink_prediction/hgnn.py`
hnhn.py	HNHN HLP pipeline on `AlgebraDataset` (negative sampling + LPE enricher)	`make run examples/hyperlink_prediction/hnhn.py`
hgnnp.py	HGNNP HLP pipeline on `AlgebraDataset` (negative sampling + LPE enricher)	`make run examples/hyperlink_prediction/hgnnp.py`
hypergcn.py	HyperGCN HLP pipeline on `AlgebraDataset` (hyperedge weights + LPE)	`make run examples/hyperlink_prediction/hypergcn.py`
mlp.py	MLP HLP pipeline on `AlgebraDataset` (negative sampling + LPE enricher)	`make run examples/hyperlink_prediction/mlp.py`
nhp.py	NHP HLP pipeline on `AlgebraDataset` (Node2Vec enricher + negative sampling)	`make run examples/hyperlink_prediction/nhp.py`
node2vecgcn.py	Compute Node2Vec embeddings then train Node2Vec+GCN HLP	`make run examples/hyperlink_prediction/node2vecgcn.py`
node2vecslp.py	Compute Node2Vec embeddings then train Node2Vec+SLP HLP	`make run examples/hyperlink_prediction/node2vecslp.py`
villain.py	VilLain HLP pipeline on `CoraDataset` (negative sampling)	`make run examples/hyperlink_prediction/villain.py`
Node classification models
common_neighbors.py	Common Neighbors NC baseline on `AlgebraDataset` (degree labels)	`make run examples/node_classification/common_neighbors.py`
gcn.py	GCN NC pipeline on `AlgebraDataset` (clique expansion + LPE enricher)	`make run examples/node_classification/gcn.py`
hgnn.py	HGNN NC pipeline on `AlgebraDataset` (LPE enricher)	`make run examples/node_classification/hgnn.py`
hnhn.py	HNHN NC pipeline on `AlgebraDataset` (LPE enricher)	`make run examples/node_classification/hnhn.py`
hgnnp.py	HGNNP NC pipeline on `AlgebraDataset` (LPE enricher)	`make run examples/node_classification/hgnnp.py`
hypergcn.py	HyperGCN NC pipeline on `AlgebraDataset` (LPE enricher)	`make run examples/node_classification/hypergcn.py`
mlp.py	MLP NC pipeline on `AlgebraDataset` (LPE enricher)	`make run examples/node_classification/mlp.py`
Sampling strategies
custom_negative_sampler.py	Custom negative sampling for HLP	`make run examples/sampling/custom_negative_sampler.py`
negative_sampler.py	Negative sampling for HLP	`make run examples/sampling/negative_sampler.py`
Splitting strategies
custom_splitter.py	Custom splitting of a dataset	`make run examples/splitting/custom_splitter.py`
Training customization
multi_model_training.py	Multi-model training for HLP	`make run examples/training/multi_model_training.py`
train_with_checkpoints.py	Training with checkpointing and prediction	`make run examples/training/train_with_checkpoints.py`
train_with_custom_logger.py	Training with a custom logger (MLP HLP, AlgebraDataset, negative sampling)	`make run examples/training/train_with_custom_logger.py`
train_with_early_stopping.py	Training with a Lightning `EarlyStopping` callback (MLP HLP, AlgebraDataset, negative sampling)	`make run examples/training/early_stopping.py`