torch_geometric (PyG) 图数据可视化

最新推荐文章于 2024-06-30 21:04:13 发布

m0_47163076

最新推荐文章于 2024-06-30 21:04:13 发布

阅读量2.2k

点赞数 2

分类专栏：图神经网络文章标签： python pytorch 深度学习

本文链接：https://blog.csdn.net/m0_47163076/article/details/120918339

版权

图神经网络专栏收录该内容

8 篇文章

订阅专栏

代码

import torch
import numpy as np
import networkx as nx
import matplotlib.pyplot as plt

def graph_showing(data):
    '''
    args:
         data: torch_geometric.data.Data
    '''
    G = nx.Graph()
    edge_index = data['edge_index'].t()
#     print(edge_index)
    edge_index = np.array(edge_index.cpu())
#     print(edge_index)
    
    G.add_edges_from(edge_index)
    nx.draw(G)
    plt.show()

展示

展示用到的是TUDataset数据集

from torch_geometric.datasets import TUDataset
dataset = TUDataset(root='./data/ENZYMES', name='ENZYMES')

for data in dataset:
    print(data)
#     print(data['edge_index'])
#     print(data.is_directed())
#     print(data.num_edges)
    graph_showing(data)