GNNE: Graph Neural Network Embeddings

GNNE: Abbreviation of Graph Neural Network Embeddings.

This repo mainly implements the mainstream Graph Embedding algorithm, which can quickly and efficiently help researchers conduct model experiments.

Installation

pip install gnne

Example

import networkx as nx
from gnne.models.node2vec import Node2Vec

# Load graph dataset
with open("datasets/example_data.txt", "r") as f:
    data = f.read()
    data = [[int(v) for v in line.split()] + [1] for line in data.split("\n")]

# Initialize graph object
graph = nx.Graph()
graph.add_weighted_edges_from(data)

# Using Node2Vec
node2vec = Node2Vec(graph, r=5, p=1, q=0.1, walk_len=10, embed_dim=2)
node2vec.train(batch_size=64, epochs=300)