set the edge attribute of complete graph in python

Question

I generate a complete graph with arbitrary nodes. Each edge has two attributes where the second one is the function of the first one. (for example, when the first one is x, the second one is 1/x). How can I add these attributes to edges?

def create_random_topology(size):     
    G = nx.generators.complete_graph(size)
    attfirst={x:random.randint(100,500) for x in G.edges()}
    nx.set_edge_attributes(G,    name="PR", values=attfirst)

    #attsecond= 1/attfirst

I want to have this graph, e.g {(0,1):{PR= 10 , BW=1/10}, (0,2):{PR=5 , BW=1/5}, (1,2):{PR=2 , BW=1/2)}} Because the graph has variable size, I won't have access to the edge with the exact name of its node, for example (1,2).

mozway · Accepted Answer · 2024-05-15 09:01:15Z

You could obviously craft another dictionary from attfirst:

nx.set_edge_attributes(G, name='BW', {e: 1/v for e, v in attfirst.items()})

Assuming this dictionary is no longer existing, you could retrieve the value from G.edges(data=True):

nx.set_edge_attributes(G, name='BW', values={tuple(e): 1/data['PR']
                                             for *e, data in G.edges(data=True)})

Example edge data (for size=5):

G.edges(data=True)

EdgeDataView([(0, 1, {'PR': 295, 'BW': 0.003389830508474576}),
              (0, 2, {'PR': 155, 'BW': 0.0064516129032258064}),
              (0, 3, {'PR': 452, 'BW': 0.0022123893805309734}),
              (0, 4, {'PR': 293, 'BW': 0.0034129692832764505}),
              (1, 2, {'PR': 169, 'BW': 0.005917159763313609}),
              (1, 3, {'PR': 204, 'BW': 0.004901960784313725}),
              (1, 4, {'PR': 487, 'BW': 0.002053388090349076}),
              (2, 3, {'PR': 322, 'BW': 0.003105590062111801}),
              (2, 4, {'PR': 440, 'BW': 0.0022727272727272726}),
              (3, 4, {'PR': 310, 'BW': 0.0032258064516129032})])

Collectives™ on Stack Overflow

set the edge attribute of complete graph in python

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related