Question

i am using the edges_iter function in networkx python…i wish to randomly extract edges from a graph and then using these edges, get the nodes at each end… i heard that edge_iter could help me with getting the random edges (i need random edges) but i cant figure out how to get the nodes attached. i’ve saved the graph in gpickle format and load it…then i use
a=G.edges_iter()
print a
and i get this
indicating the occurance of an edge…what i want next is to find out the nodes that are attached at the ends of this edge…help me out with this…thanks a lot guys.

Answer 1

edges_iter returns a generator object, so I guess you want to iterate through the edges that you get.

Anyway, the edges already contain information about the nodes they are associated with, since each edge can be seen as a tuple of two nodes. Consider this example:

import networkx as nx

g = Graph()
g.add_edges_from([(1, 2), (2, 3), (4,5), (1, 5), (2, 4)])
# Nodes are automatically added

for edge in g.edges_iter():
    print edge
    print 'Node 1:', edge[0]
    print 'Node 2:', edge[1]

edge[0] will be one node and edge[1] will be the second.

However, I don’t see how you would get random edges with this approach. Another solution to do that could be to use random.choice and just select edges with that:

from random import choice

... # Initialize graph

random_edge = choice(g.edges())

If you instead want to select a number of random edges, you could use random.sample:

from random import sample

... # Initialize graph

# Get 2 random edges
random_edges = sample(g.edges(), 2)