Adding randomized attributes to graph edges

January 7, 2022

I’ve tried to add, randomly, one of two values of attribute in a dict, to each edge of some listed graphs, through the following code:

import networkx as nx
import random

Glist = []

for _ in range(12):
    g = nx.erdos_renyi_graph(n = 20, p = random.random())
    Glist.append(g)

for i in range(len(Glist)):
    for u, v in Glist[i].edges():
        attribs = {Glist[i].edges(): {'relation': random.choice(['friend', 'enemy'])}}

def set_Net_att(my_list, my_dict):
    for _ in my_list:
        for i in range(len(my_list)):
            gatt = nx.set_edge_attributes(my_list[i], my_dict)

set_Net_att(Glist, attribs)

print(Glist[2].edges(data = True))

But I got this error:

TypeError                                 Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_14600/893381260.py in <module>
     14 for i in range(len(Glist)):
     15     for u, v in Glist[i].edges():
---> 16         attribs = {Glist[i].edges(): {'relation': random.choice(['friend', 'enemy'])}}
     17 
     18 def set_Net_att(my_list, my_dict):

TypeError: unhashable type: 'EdgeView'

How can I get things done? I’d really appreciate your support.

>Solution :

You do not construct the dictionary for the function set_edge_attributes correctly (and, in general, use too many non-Pythonic features in your code). Here is a correct (and improved) solution:

# Build a list of graphs
glist = [nx.erdos_renyi_graph(n = 20, p = random.random()) for _ in range(12)]

# Generate and assign attributes
for g in glist:
    attribs = {edge: random.choice(['friend', 'enemy']) for edge in g.edges}
    nx.set_edge_attributes(g, attribs, "relation")