def betweenness_centrality_labeling(self,graph,approx=None):
result={}
labeled_graph=nx.Graph(graph)
if approx is None:
centrality=list(nx.betweenness_centrality(graph).items())
else:
centrality=list(nx.betweenness_centrality(graph,k=approx).items())
sorted_centrality=sorted(centrality,key=lambda n:n[1],reverse=True)
dict_={}
label=0
for t in sorted_centrality:
dict_[t[0]]=label
label+=1
nx.set_node_attributes(labeled_graph,dict_,'labeling')
ordered_nodes=list(zip(*sorted_centrality))[0]
result['labeled_graph']=labeled_graph
result['sorted_centrality']=sorted_centrality
result['ordered_nodes']=ordered_nodes
return result
def addCentralities(self):
"""Add centrality coeficients for the final version of the graph.
Centralities include : Kalz centrality, global clustering,
square clustering, closeness centrality, harmonic centrality,
betweenness centrality.
From all centralities we choose the max or the mean. More about
these centralities can be found in networkX algorithms.
"""
self.kalz_coef = max(list(nx.katz_centrality(self.graph).values()))
self.glob_clust_coef = mean(list(nx.clustering(self.graph).values()))
self.square_clustering_coef = mean(
list(nx.square_clustering(self.graph).values()))
self.harmonic_coef = max(
list(nx.harmonic_centrality(self.graph).values()))
self.betweenness_coef = max(
list(nx.betweenness_centrality(self.graph).values()))
self.closeness_coef = max(
list(nx.closeness_centrality(self.graph).values()))
def _update_stat_json(self, stat_name):
"""
Callback function for updating the statistic shown.
:param stat_name: (str) Name of the statistic to display (e.g. graph_summary).
:return: (json) Json of the graph information depending on chosen statistic.
"""
switcher = {
"graph_summary":
self.get_graph_summary(),
"average_degree_connectivity":
nx.average_degree_connectivity(self.graph.get_graph()),
"average_neighbor_degree":
nx.average_neighbor_degree(self.graph.get_graph()),
"betweenness_centrality":
nx.betweenness_centrality(self.graph.get_graph()),
}
if type(self.graph).__name__ == "PMFG":
switcher["disparity_measure"] = self.graph.get_disparity_measure()
return json.dumps(switcher.get(stat_name), indent=2)
def graph_properties(G, n, p):
# graph properties
A = nx.adjacency_matrix(G).toarray()
L = nx.laplacian_matrix(G).toarray()
betwn_ctrlity = nx.betweenness_centrality(G)
names = ['id', 'data']
formats = ['i', 'f8']
dtype = dict(names=names, formats=formats)
betwn_ctrlity_array = np.array(list(betwn_ctrlity.items()), dtype=dtype)
c = nx.average_clustering(G)
d = nx.diameter(G)
np.savetxt('Q8_Assets/Adjacency_N_' + str(n) + '_P_' + str(p) + '.csv',
A,
delimiter=",")
np.savetxt('Q8_Assets/Laplacian_N_' + str(n) + '_P_' + str(p) + '.csv',
L,
delimiter=",")
np.savetxt('Q8_Assets/Betweeness_N_' + str(n) + '_P_' + str(p) + '.csv',
betwn_ctrlity_array,
delimiter=",")
print('Clustering = ', c, '\n')
print('Diameter = ', d, '\n')
def calculate(self, graph, params):
nx_graph = util.to_networkx(graph)
return nx.betweenness_centrality(nx_graph)
def calculate(self, graph, params):
nx_graph = util.to_networkx(graph)
return nx.betweenness_centrality(nx_graph)
__author__ = 'nourl'
from common.transform import *
from PyCommDete import *
from networkx import nx
from inputs.formal_edgelist import *
from multiprocessing import Pool
from sys import exit
#C = nx.read_gml(filelist[1])
C = nx.Graph(formal_edgelist('./benchmarks/network.dat'))
print "length:", len(C)
betw_ori = nx.betweenness_centrality(C)
degr_ori = C.degree()
nodes = C.nodes()
betw_ori = sorted(betw_ori.iteritems(), key=lambda x: x[1], reverse=True)
average = sum(x[1] for x in betw_ori) / len(betw_ori)
betw = [x for x in betw_ori if x[1] >= 0.2 * average]
betw_ex_nei = [x[0] for x in betw]
bitmap = [0] * len(betw_ori)
print "len of betw_ex_nei:", len(betw_ex_nei)
for x in betw_ex_nei:
bitmap[x - 1] = x
recover = []
for x in betw_ex_nei:
if bitmap[x - 1] > 0:
recover.append(x)
nei = C.neighbors(x)
__author__ = 'nourl'
from common.transform import *
from PyCommDete import *
from networkx import nx
from inputs.formal_edgelist import *
from multiprocessing import Pool
from sys import exit
#C = nx.read_gml(filelist[1])
C = nx.Graph(formal_edgelist('./benchmarks/network.dat'))
print "length:",len(C)
betw_ori = nx.betweenness_centrality(C)
degr_ori = C.degree()
nodes = C.nodes()
betw_ori = sorted(betw_ori.iteritems(), key=lambda x:x[1],reverse=True)
average = sum(x[1] for x in betw_ori)/len(betw_ori)
betw=[x for x in betw_ori if x[1]>=0.2*average]
betw_ex_nei=[x[0] for x in betw]
bitmap = [0]*len(betw_ori)
print "len of betw_ex_nei:",len(betw_ex_nei)
for x in betw_ex_nei:
bitmap[x-1]=x
recover=[]
for x in betw_ex_nei:
if bitmap[x-1] >0:
recover.append(x)
nei = C.neighbors(x)
def get_all_nodes(netw,seeds_type):
if seeds_type == 1:
orig = C.degree()
elif seeds_type == 2:
orig = nx.betweenness_centrality(netw)
elif seeds_type == 3:
betw = nx.betweenness_centrality(netw)
degr = netw.degree()
nodes = netw.nodes()
betw = sorted(betw.iteritems(), key=lambda x:x[1],reverse=True)
bitmap = [0]*len(betw)
betw_ex_nei=[x[0] for x in betw]
print "len of betw_ex_nei:",len(betw_ex_nei)
for x in betw_ex_nei:
bitmap[x-1]=x
recover=[]
for x in betw_ex_nei:
if bitmap[x-1] >0:
recover.append(x)
nei = C.neighbors(x)
for n in nei:
bitmap[n-1] = 0
for x in recover:
bitmap[x-1]=x
seed_betw = [node for node in nodes if bitmap[node-1]]
print "len of betw:",len(seed_betw)
print "seed_betw:",seed_betw
degr = sorted(degr.iteritems(), key=lambda x:x[1],reverse=True)
bitmap = [0]*len(degr)
degr_ex_nei=[x[0] for x in degr]
print "len of degr_ex_nei:",len(degr_ex_nei)
for x in degr_ex_nei:
bitmap[x-1]=x
recover=[]
for x in degr_ex_nei:
if bitmap[x-1] >0:
recover.append(x)
nei = C.neighbors(x)
for n in nei:
bitmap[n-1] = 0
for x in recover:
bitmap[x-1]=x
seed_degr = [node for node in nodes if bitmap[node-1]]
print "len of degr:",len(seed_degr)
print "seed_degr",seed_degr
seed_cross=list(set(seed_betw).intersection(set(seed_degr)))
print "len of seed_cross:",len(seed_cross)
print "seed_cross:",seed_cross
return seed_cross
nodes = netw.nodes()
orig=sorted(orig.iteritems(), key=lambda x:x[1],reverse=True)
bitmap = [0]*len(orig)
average = sum(x[1] for x in orig)/len(orig)
orig=[x for x in orig if x[1]>=average]
orig_over_ave=[x[0] for x in orig]
for x in orig_over_ave:
bitmap[x-1]=x
recover=[]
for x in orig_over_ave:
if bitmap[x-1] >0:
recover.append(x)
nei = C.neighbors(x)
for n in nei:
bitmap[n-1] = 0
for x in recover:
bitmap[x-1]=x
seed = [node for node in nodes if bitmap[node-1]]
return seed
