Sort multi edge nodes with networkx in python

import

import networkx as nx
import numpy as np

Create multi graph

G = nx.MultiGraph()
G.add_nodes_from(["A", "B", "C", "D", "E", "F"])
G.add_edges_from([("A", "B"), ("B", "C"), ("B", "D"), ("C", "D"), ("A", "E"), ("C", "E"), ("C", "F"), ("C", "F"), ("F", "F")])

• this graph has multi edge in between node C and F
• in this case, we want to visualize its multi edges in matplotlib

Approach1. Convert data structure G to Adjacency matrix

• "Adjacency matrix" is the way to describe network data structure by using matrix.
• It is described as below

A = (nx.adjacency_matrix(G)).todense()

// output of A
matrix([[0., 1., 0., 0., 1., 0.],
[1., 0., 1., 1., 0., 0.],
[0., 1., 0., 1., 1., 2.],
[0., 1., 1., 0., 0., 0.],
[1., 0., 1., 0., 0., 0.],
[0., 0., 2., 0., 0., 1.]])

• next, we want to sort its matrix with the filter.
• The filter serves some nodes which have multi edges.
• How to generate the filter
• Flatten A to one dimension array or list
• get index number from it by sorting
• get node value from origin node list by using above index number

node_list = list(G)
A_flatten = np.ravel(A)
idx_sorted_list = np.where(A_flatten >= 2)[0] # return value is tuple
[node_list[(I % len(G.nodes))] for I in idx_sorted_list]

//
['F', 'C']

Then after, let's try to apply this way to existed dataset. In this time, we use zachary's karate club graph.

source code  

import networkx as nx
import numpy as np
import matplotlib.pyplot as plt

def get_multi_edge_nodes(G, threshold):
    node_list = list(G)
    A = nx.adjacency_matrix(G)
    A_flatten = np.ravel(A.todense())
    idx_sorted_list = np.where(A_flatten >= threshold)[0] # return value is tuple
    node_multi_edge_list = [node_list[(I % len(G.nodes))] for I in idx_sorted_list]
    return set(node_multi_edge_list)


def rand_edges(node_num, edge_num):
    def randint(low, high, size):
        return np.random.randint(low, high, size)
    return [(s, t) for s, t in zip(randint(0, node_num, edge_num), randint(0, node_num, edge_num))]

 

G_01 = nx.karate_club_graph()
G_02 = nx.MultiGraph()
G_02.add_edges_from(G_01.edges)
G_02.add_edges_from(rand_edges(len(G_02), 20))

 

ret1 = get_multi_edge_nodes(G_01, 2)
ret2 = get_multi_edge_nodes(G_02, 2)

 

G_02_sub = G_02.subgraph(ret2)

 

fig, (ax1, ax2) = plt.subplots(1, 2)
nx.draw_networkx(G_01, ax=ax1)
nx.draw_networkx(G_02_sub, ax=ax2)

plt.savefig('test.png')