此问题已在此处有答案：

how to draw directed graphs using networkx in python?（6个回答）
上个月关门了。
我刚刚写了python 3.4代码，用于使用Networkx创建DAG（有向无环图）图。
问题是我无法从内部访问字典值：

nx.draw_networkx_nodes(dag,
                       pos,
                       node_color=[node[1]['color'] for node in dag.nodes(data=True)], 
                       node_shape=[node[1]['shape'] for node in dag.nodes(data=True)])

下面是我的代码：

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

file_path = 'Demo.xlsx'

df = pd.read_excel(file_path)

dag = nx.DiGraph()

for col in treatment_columns:
    dag.add_node(col, label='Treatment', color='blue', shape='ellipse')

for col in confounder_columns:
    dag.add_node(col, label='Confounder', color='green', shape='rectangle')

dag.add_node(outcome_column, label='Outcome', color='red', shape='ellipse')

pos = nx.spring_layout(dag)

nx.draw_networkx_nodes(dag, pos, node_color=[node[1]['color'] for node in dag.nodes(data=True)], node_shape=[node[1]['shape'] for node in dag.nodes(data=True)])
nx.draw_networkx_labels(dag, pos, labels={node[0]: node[0] for node in dag.nodes(data=True)})
nx.draw_networkx_edges(dag, pos)

plt.axis('off')
plt.title("Causal DAG")
plt.show()