fix information loss in undirected graph

[Ankush-Chander]

In an undirected graph, when we add duplicate edges, the associated weight is overwritten instead of accumulated.
As per networkx documentation:

Adding the same edge twice has no effect but any edge data will be updated when each duplicate edge is added.

For a hypothetical example:

weighted_edges = [
    ("play", "with", {'weight': 2.0}),
    ("with", "play", {'weight': 11.0}),
]

graph= nx.Graph()
graph.add_edges_from(weighted_edges)

current code leads to(overwriting):

("play", "with", 11.0)

Desired behaviour(accumulation):

("play", "with", 13.0)

Thanks to @yamika-g for pointing it out here: