What is the most efficient way to move/rename a node in NetworkX?

Question

I am using the NetworkX graph library for Python. At some point in my program I would like to "consolidate" my nodeIDs into a sequence of numbers. Here's my naive approach:

start = 1 # could be anything
for i, n in enumerate(g.nodes()):
    if i+start == n:
        continue
    g.add_node(i+start, attr_dict=g.node[n])
    g.add_edges_from([(i+start, v, g[n][v]) for v in g.neighbors(n)])
    g.remove_node(n)

Is there a faster way than this exhaustive copy of all the neighbors? For example, I tried g[i+start] = g[n], but that is forbidden.

Thanks!

Answer 1

Would this work?

http://networkx.github.io/documentation/latest/reference/generated/networkx.relabel.convert_node_labels_to_integers.html

import networkx as nx
G = nx.Graph()
G.add_node(1)
G.add_nodes_from('spam')
print G.nodes()

returns:

['a', 1, 's', 'm', 'p']

now:

start = 1
G = nx.convert_node_labels_to_integers(G,first_label=start)
print G.nodes()

returns:

[1, 2, 3, 4, 5]

Answer 2

In case your interest is still relevant, there is networkx.relabel_nodes() which takes a mapping dictionary.