Package index

Layers

Graph neural network layer implementations

layer_gcn()

GCN Convolutional Layer (Kipf & Welling 2016)

layer_gcn_general()

Generalized GCN Layer (Hamilton 2020)

layer_sage()

GraphSAGE Layer (Hamilton et al. 2017)

layer_gat()

Graph Attention Network Layer (Veličković et al. 2018)

layer_gin()

Graph Isomorphism Network Layer (Xu et al. 2019)

layer_regconv()

Regional GCN Convolutional Layer (Guo et al. 2025)

Models

Pre-built graph neural network models

model_gcn()

Multi-layer GCN Model

model_gcn_general()

Multi-layer Generalized GCN Model (Hamilton 2020)

model_sage()

Multi-layer GraphSAGE Model (Hamilton et al. 2017)

model_gat()

Multi-layer Graph Attention Network Model (Veličković et al. 2018)

model_gin()

Multi-layer Graph Isomorphism Network Model (Xu et al. 2019)

Graph Utilities

Functions for constructing and manipulating graphs

graph_split()

Create Train/Validation/Test Split for Graph Data

adj_from_edgelist()

Create Sparse Adjacency Matrix from Edge List

nodes_to_tensor()

Convert Node Features to Tensor

gcn_normalize() adj_row_normalize() add_graph_self_loops()

Add self-loops to a graph

Pooling

Global pooling operations for graph-level representations

pool_global_add() pool_global_mean() pool_global_max()

Graph Pooling Functions

Aggregators

Message passing aggregators for combining neighbor features

Aggregator() SumAggregator() MeanAggregator() MaxAggregator() MinAggregator() ProductAggregator() VarAggregator() StdAggregator() LSTMAggregator() SoftmaxAggregator()

Message Passing Aggregators