过滤NetworkX图形以列出来自具有特定属性的节点的所有边

Question

I have a NetworkX graph, with various node attributes loaded into the graph. I would like to iterate through this graph, and first find all of the nodes that contain the attribute 'metric' : 'Top2' . Then once I have this list, I would like to the edges of all the 'Top2' nodes, and list out any edge that a 'Top2' node has to a node with the attribute 'city': 'Dallas' '

I am able to successfully load my graph and attributes into NetworkX, but I can't seem to figure out how to get the node attribute filtering part figured out like I want to.

Ultimately, I want to find any connections where a node with the attribute 'Top2' is in direct contact with a node with the attribute 'Dallas':

Code

# import module
import networkx as nx

# create a blank graph
g = nx.Graph()

# add nodes with attributes to the graph
n = [(u'A', {'city': 'Dallas'}),
     (u'C', {'city': 'Toledo'}),
     (u'B', {'city': 'Dallas'}),
     (u'E', {'city': 'Toldeo'}),
     (u'D', {'city': 'Lousiville', 'metric': 'Top2'}),
     (u'G', {'city': 'Lousiville'}),
     (u'F', {'city': 'Dallas', 'metric': 'Top2'})]

e = [(u'A', u'B'),
     (u'C', u'B'),
     (u'B', u'D'),
     (u'E', u'D'),
     (u'E', u'G'),
     (u'E', u'F'),
     (u'D', u'G'),
     (u'D', u'F'),
     (u'G', u'F')]

# add nodes and edges to the graph
g.add_nodes_from(n)
g.add_edges_from(e)

# draw the graph
nx.draw(g)

# iterate through graph to identify all nodes with attribute 'metric' == 'Top2'
Top2_nodes = []
for n in g.nodes(data=True) if n[1]['metric'] == 'Top2'
    Top2_nodes.append(n)

# iterate through 'Top2_nodes' to identify their connections to nodes with attribute 'city': 'Dallas'
...

Ideally, the results would be formatted as follows:

D,B
D,F
D,G

Answer 1

This uses 2 helper-functions that return a bool. Your actual result is returned as list comprehension in the end.

nodes = g.node

def has_top2(n):
    return 'metric' in nodes[n] and nodes[n]['metric'] == 'Top2'

def has_dallas(n):
    return 'city' in nodes[n] and nodes[n]['city'] == 'Dallas'

[(n, u) for n, u in g.edges() 
    if (has_top2(n) and has_dallas(u)) 
    or (has_top2(u) and has_dallas(n))
]