In [1]:

    

from networkit import *


    

In [2]:

    

# we find the k nodes with highest closeness in a graph
G = readGraph("input/PGPgiantcompo.graph", Format.METIS)


    

In [3]:

    

k = 10


    

In [12]:

    

# The last two parameters indicate which heuristics we want to use.
# We recommend using True, False for complex networks and True, True for street networks
topk = centrality.TopCloseness(G, k, True, False)
%time topk.run()


    

    




CPU times: user 1.64 s, sys: 25.4 ms, total: 1.67 s
Wall time: 493 ms






    
Out[12]:





<_NetworKit.TopCloseness at 0x10ffc2898>

In [8]:

    

cc = centrality.Closeness(G)
%time cc.run()


    

    




CPU times: user 1min 19s, sys: 600 ms, total: 1min 19s
Wall time: 22.4 s






    
Out[8]:





<_NetworKit.Closeness at 0x10ffc2550>

In [13]:

    

topk.topkNodesList()


    

    
Out[13]:





[1143, 6655, 6555, 1435, 7297, 6859, 6744, 6768, 6098, 2258]

In [14]:

    

cc.ranking()[:k]


    

    
Out[14]:





[(1143, 2.1164469089292894e-05),
 (6655, 2.0922252908193154e-05),
 (6555, 2.0463707614545603e-05),
 (1435, 2.0415246105791806e-05),
 (7297, 2.0345466012899025e-05),
 (6859, 2.0323550930818634e-05),
 (6744, 2.0264245764772636e-05),
 (6768, 1.9864129355210362e-05),
 (6098, 1.9855452307203558e-05),
 (2258, 1.9812573058863156e-05)]

In [ ]: