如何基于DataFrame构建权重>0.2的NetworkX有向图？

首先明确你的DataFrame结构，下面分两种常见场景给出完整实现：

场景1：DataFrame是邻接矩阵格式（行和列都是节点，单元格值为权重） #

假设你的DataFrame df 的行索引和列名都是节点标识，单元格值是对应节点间的权重：

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

G = nx.DiGraph() 

# 添加节点：用行索引作为节点标识
nodes = df.index.tolist()
G.add_nodes_from(nodes)

# 添加符合条件的边：遍历所有节点对，权重>0.2则添加
for u in nodes:
    for v in nodes:
        weight = df.loc[u, v]
        if weight > 0.2 and u != v:  # 可根据需求去掉u!=v保留自环
            G.add_edge(u, v, weight=weight)

# 绘图执行
pos = nx.spring_layout(G)
nx.draw_networkx_nodes(G, pos, node_size=100)
nx.draw_networkx_labels(G, pos, font_size=10, font_family='sans-serif')
edges = G.edges()
weights = [G[u][v]['weight'] for u, v in edges]
nx.draw_networkx_edges(G, pos, edgelist=edges, width=weights)
plt.axis('off')
plt.show()

场景2：DataFrame是长格式（包含from_node、to_node、weight列） #

如果你的DataFrame包含出发节点、目标节点、权重三列：

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

G = nx.DiGraph() 

# 添加节点：合并出发/目标节点的唯一值
nodes = pd.concat([df['from_node'], df['to_node']]).unique().tolist()
G.add_nodes_from(nodes)

# 添加符合条件的边：筛选权重>0.2的行批量添加
filtered_edges = df[df['weight'] > 0.2]
for _, row in filtered_edges.iterrows():
    G.add_edge(row['from_node'], row['to_node'], weight=row['weight'])

# 绘图执行（和上方一致）
pos = nx.spring_layout(G)
nx.draw_networkx_nodes(G, pos, node_size=100)
nx.draw_networkx_labels(G, pos, font_size=10, font_family='sans-serif')
edges = G.edges()
weights = [G[u][v]['weight'] for u, v in edges]
nx.draw_networkx_edges(G, pos, edgelist=edges, width=weights)
plt.axis('off')
plt.show()

关键注意点： #

• 确保节点标识在DataFrame中格式统一（均为字符串或数字）
• 添加边时必须传入weight属性，否则后续提取权重会报错
• 可根据需求选择是否保留节点自环

内容的提问来源于stack exchange，提问作者sylvia_zhang