【NetworkX】Graph Optimization with NetworkX in Python

NetworkX 图优化算例：『Intro to Graph Optimization with NetworkX in Python - DataCamp』

使用 networkx 版本为 2.4 版。

---

【1】g.node[nlrow['id']] = nlrow[1:].to_dict() 

报错 AttributeError: 'Graph' object has no attribute 'node'

新版改为 g.nodes[nlrow['id']].update(nlrow[1:].to_dict()) 

【2】 g.edges(data=True)[0:5] 

报错 TypeError: 'EdgeDataView' object is not subscriptable

新版改为 list(g.edges(data=True))[0:5] 

同理， g.nodes(data=True)[0:10] 改为 list(g.nodes(data=True))[0:10] 

【3】 edge_colors = [e[2]['color'] for e in g.edges(data=True)] 

报错 KeyError: 'color'

尝试查看 [e[2] for e in g.edges(data=True)] ，输出为：

[{'attr_dict': {'trail': 'rs',
   'distance': 0.3,
   'color': 'red',
   'estimate': 0}},
 {'attr_dict': {'trail': 'rs',
   'distance': 0.21,
   'color': 'red',
   'estimate': 0}},
...
]

因此改为 [e[2]['attr_dict']['color'] for e in g.edges(data=True)] 

【4】 nodes_odd_degree = [v for v, d in g.degree_iter() if d % 2 == 1] 

报错 AttributeError: 'Graph' object has no attribute 'degree_iter'

新版改为 nodes_odd_degree = [v for v, d in g.degree() if d % 2 == 1] 

【5】 odd_matching = list(pd.unique([tuple(sorted([k, v])) for k, v in odd_matching_dupes.items()])) 

报错 AttributeError: 'set' object has no attribute 'items'

改为 odd_matching = list(pd.unique([tuple(sorted([k, v])) for k, v in odd_matching_dupes]))