NetworkX绘制multigraph时无法显示边标签如何解决

方案1：使用NetworkX原生绘图实现边标签显示 #

原代码的核心问题有两个：

1. 构造图时未将relation列作为边属性存储，导致后续无法读取边标签
2. 没有调用专门的边标签绘制接口

修正后的完整可运行代码如下：

import pandas as pd
import networkx as nx
import matplotlib.pyplot as plt

minidf = pd.DataFrame(data={'relation': ['subject', 'subject', 'broader'], 
                       'source': ['pmt3423', 'pmt2040', 'category:myoblasts'], 
                       'target': ['conceito', 'frio', 'category:non-terminally_differentiated_(blast)']})
# 新增edge_attr参数，将relation存入边属性
miniG = nx.from_pandas_edgelist(minidf,'source', 'target',
                      edge_key='relation', edge_attr='relation', create_using=nx.MultiDiGraph())

# 固定布局坐标，保证节点和边标签位置对齐
pos = nx.spring_layout(miniG, seed=42)
# 绘制节点、边、节点标签
nx.draw_networkx(miniG, pos=pos, with_labels=True, node_size=3000, font_size=8)
# 读取边的relation属性作为标签
edge_labels = nx.get_edge_attributes(miniG, 'relation')
# 绘制边标签
nx.draw_networkx_edge_labels(miniG, pos=pos, edge_labels=edge_labels, font_size=7)

plt.axis('off')
plt.show()

方案2：使用Pyvis生成带边标签的交互知识图谱 #

Pyvis生成的网页版图谱支持缩放、拖拽节点，更适合知识图谱浏览，正确配置代码如下：

import pandas as pd
import networkx as nx
from pyvis.network import Network

minidf = pd.DataFrame(data={'relation': ['subject', 'subject', 'broader'], 
                       'source': ['pmt3423', 'pmt2040', 'category:myoblasts'], 
                       'target': ['conceito', 'frio', 'category:non-terminally_differentiated_(blast)']})
miniG = nx.from_pandas_edgelist(minidf,'source', 'target',
                      edge_key='relation', edge_attr='relation', create_using=nx.MultiDiGraph())

# 初始化有向图，非Jupyter环境可将notebook参数改为False
net = Network(directed=True, notebook=True, height='800px', width='100%')
net.from_nx(miniG)
# 给所有边设置显示标签
for edge in net.edges:
    edge['label'] = edge['relation']
    edge['font.size'] = 10

# 生成HTML文件，打开即可查看带边标签的交互图谱
net.show('knowledge_graph.html')

方案3：Graphviz+neato渲染带边标签的静态图 #

之前生成dot文件无标签是因为未给边设置label属性，修正代码如下：

import pandas as pd
import networkx as nx

minidf = pd.DataFrame(data={'relation': ['subject', 'subject', 'broader'], 
                       'source': ['pmt3423', 'pmt2040', 'category:myoblasts'], 
                       'target': ['conceito', 'frio', 'category:non-terminally_differentiated_(blast)']})
miniG = nx.from_pandas_edgelist(minidf,'source', 'target',
                      edge_key='relation', edge_attr='relation', create_using=nx.MultiDiGraph())

# 给边添加label属性，导出dot时会自动识别
for u, v, data in miniG.edges(data=True):
    data['label'] = data['relation']

# 导出dot文件
nx.nx_pydot.write_dot(miniG, 'kg.dot')

导出dot文件后执行命令渲染即可得到带边标签的图片：
neato -Tpng kg.dot -o kg.png

内容的提问来源于stack exchange，提问作者DSchlingel