Querying undirected edge objects in an undirected networkx graph - networkx

Question

Given an undirected graph in networkx:
graph=Graph()
graph.add_edge(1,2)
graph.add_edge(2,3)
When querying the edgeset of this graph using the method graph.edges(),the result is returned as a tuple, i.e. a directed view of the edgeset. In the example above, graph.edges() returns:
[(1, 2), (2, 3)]
If I would query all edges incident to vertex 2 (graph.edges(2)) I get a different directed view:
[(2, 1), (2, 3)]
This behavior is problematic if I have for instance an external dict that maps an edge (key) to some other object. For instance:
lookup={edge : some_object(edge) for edge in graph.edges()}
for edge in graph.edges(2):
print(lookup(edge)) #throws key not found exception
The exception is thrown because key (1,2) exists, but (2,1) does not. What would be the best (efficient/pythonic/...) way to get an undirected object when iterating over an undirected graph in networkx? I could associate some_object(edge) with an edge as one of its attributes, but that doesn't really answer the question.

Answer 1

dfs_labeled_edges(G, source) Traverses through the graph using depth first search.
Since dfs_edges stops after reaching every node, it will not always detect every edge. Instead, dfs_labeled_edges can be implemented to obtain all the edge objects in an undirected graph.
Code:
import networkx as nx
from networkx import dfs_labeled_edges
nodes = [1,2,3]
edges = [(1,2), (2,3)]
G = nx.Graph()
G.add_nodes_from(nodes)
G.add_edges_from(edges)
undirectedGraphEdges = []
dfs = list(dfs_labeled_edges(G,1))
for edge in dfs:
if(edge[:2] not in undirectedGraphEdges):
undirectedGraphEdges.append(edge[:2])
if undirectedGraphEdges[0] not in edges:
del undirectedGraphEdges[0]
print(undirectedGraphEdges)
# [(1, 2), (2, 1), (2, 3), (3, 2)]