CODEWITHSUNDEEP
pythonnetworkx
Hryhorii Pavlenko
Hryhorii Pavlenko

Reputation: 3910

How do I create a directed graph in networkx from a dictionary?

I have ~10k publications that have inbound or/and outbound citations.

Data comes in the following format (two entries as an example):

# each 'number' is a 'paper_id'
citations = {
    '157553241': { 
        'inbound_citations': [],
        'outbound_citations': [
            '141919793',
            '158546657',
            '156580052',
            '159778536',
            '157021328',
            '158546657',
            '157021328',
            '141919793',
            '153005744',
            '159778536',
            '112335878',
            '156580052'
        ]
    },
    '54196724': {
        'inbound_citations': ['204753337', '55910675'],
        'outbound_citations': ['153776751', '141060228', '33718066', '158233543']
    },
}

How do I transform this format into something I could feed to networkx?

I'd like to find the most 'central' papers & discover some cliques (to begin with).

I've tried

G = nx.DiGraph(citations)

but I don't think it works like that...

Upvotes: 1

Views: 1443

Answers (1)

Corralien
Corralien

Reputation: 120559

You need to build a list of edges like this:

import networkx as nx
import matplotlib.pyplot as plt

edges = []
for node in citations:
    for parent in citations[node]['inbound_citations']:
        edges.append((parent, node))
    for child in citations[node]['outbound_citations']:
        edges.append((node, child))

G = nx.DiGraph()
G.add_edges_from(edges)

nx.draw(G, with_labels=True)
plt.show()

enter image description here

Upvotes: 2

Related Questions