def calc_cluster_coefficient(hcc, V):
# check max weight first (random groups may have a max of 0)
max_w = max([w for u, v, w in hcc.edges(data='weight')])
if max_w == 0:
return 0
else:
return sum(clustering(hcc, weight='weight').values()) / float(len(V)) #average_clustering(hcc)
def get_topological_features(G, nodes=None):
N_ = len(G.nodes)
if nodes is None:
nodes = G.nodes
# Degree centrality
d_c = get_features(degree_centrality(G).values())
print 'a'
# Betweeness centrality
b_c = get_features(betweenness_centrality(G).values())
print 'b'
# Close ness centrality
c_c = get_features(closeness_centrality(G).values())
print 'c'
# Clustering
c = get_features(clustering(G).values())
print 'd'
d = diameter(G)
r = radius(G)
s_p_average = []
for s in shortest_path_length(G):
dic = s[1]
lengths = dic.values()
s_p_average += [sum(lengths) / float(N_)]
s_p_average = get_features(s_p_average)
features = np.concatenate((d_c, b_c, c_c, c, s_p_average, [d], [r]),
axis=0)
return features
def clustering_coefficient_and_path_length(G):
clustering_values = clustering(G)
clustering_values = np.mean(
[abs(value) for value in clustering_values.values()])
path_length = []
for C in (G.subgraph(c).copy() for c in nx.connected_components(G)):
path_length.append(average_shortest_path_length(C))
path_length = np.mean(path_length)
return clustering_values, path_length
def attack_degree(graph, nodes_number, iteration_removals):
after_removals = []
lenght_range = int((nodes_number - iteration_removals)/iteration_removals)
degress = list(graph.degree())
degress.sort(key=lambda x: x[1], reverse=True)
first_element = first_tuple_element(degress)
degress_reshaped = np.reshape(
first_element, (iteration_removals, int(nodes_number/iteration_removals))).T
for iterator in tqdm(range(lenght_range)):
# remove iterations_removals do maior componente conexo da rede
graph.remove_nodes_from(degress_reshaped[iterator])
after_removals.append(
np.mean(list(nx_cluster.clustering(graph).values())))
return after_removals
def clustering_coefficient(self, use_undirected=False):
if use_undirected:
g = self.g.to_undirected()
else:
g = self.g
if not self.department_clusters:
self.extract_departments()
res = []
for department, cluster in self.department_clusters.items():
cfs = clustering(g, cluster)
res.append((department, sum(cfs.values()) / len(cfs)))
res = sorted(res, key=lambda x: x[1], reverse=True)[:self.top_k]
return res
def __graph_props(self):
if self.__data_dict["Connected"]:
self.__data_dict['Diameter'] = nx.diameter(self.__g)
self.__data_dict['Radius'] = nx.radius(self.__g)
self.__data_dict["Centered nodes"] = list(nx.center(self.__g))
else:
self.__data_dict['Diameter'] = None
self.__data_dict['Radius'] = None
self.__data_dict["Centered nodes"] = []
if self.__data_dict["Type"] == "Undirected":
self.__data_dict['Average Clustering'] = nx.average_clustering(
self.__g)
self.__clustering_coefficients = cluster.clustering(self.__g)
else:
self.__data_dict['Average Clustering'] = None
def random_attack(graph, nodes_number, iteration_removals):
# lista onde estão os valores de clusterização da rede depois da remoção dos nos
after_removals = []
# number_nodes_largest_cc = [] #número de nós do maior componente da rede
lenght_range = int((nodes_number - iteration_removals)/iteration_removals)
for _ in tqdm(range(lenght_range)):
# retorna um set, com os nos que representam o maior componente conexo da rede
largest_cc = max(networkx.connected_components(graph), key=len)
# se o tamanho do maior componente conexo por menor que o número de nos que quero remover, sair do loop
if len(largest_cc) < iteration_removals:
break
# number_nodes_largest_cc.append(len(largest_cc))
# remove iterations_removals do maior componente conexo da rede
graph.remove_nodes_from(random.sample(largest_cc, iteration_removals))
after_removals.append(
np.mean(list(nx_cluster.clustering(graph).values())))
return after_removals
def nx_average_clustering_per_k(g):
coefficients = [None] * g.number_of_nodes()
for k in range(len(coefficients)):
coefficients[k] = [0,0]
clustering_coefficient = clustering(g) # dict of (vertex, cc)
all_degrees = g.degree() # list of (vertex, degree)
for deg in all_degrees:
coefficients[deg[1]][0] += 1
coefficients[deg[1]][1] += clustering_coefficient[deg[0]]
# average cc
ck = []
for coef in coefficients:
if coef[0] == 0:
ck.append(0)
else:
ck.append(coef[1]/coef[0])
return ck
def clustering_coefficient(g,
department_clusters=None,
use_undirected=False,
k=20):
if use_undirected:
g = g.to_undirected()
else:
g = g
if not department_clusters:
department_clusters = extract_departments(g)
res = []
for department, cluster in department_clusters.items():
cfs = clustering(g, cluster)
res.append((department, sum(cfs.values()) / len(cfs)))
res = sorted(res, key=lambda x: x[1], reverse=True)[:k]
return res
def graph_stats(G):
"""
Compute all the graph-related statistics in the features.
Note that since the graph is always fully connected, all of these are the
weighted versions. For this reason, many of these functions use the
implementations in bctpy rather than NetworkX.
"""
# Local measures
clustering_dict = clustering(G, weight='weight')
adjacency = np.array(adjacency_matrix(G).todense())
betweenness_centrality_dict = betweenness_centrality(G, weight='weight')
paths = shortest_path_length(G, weight='weight')
eccentricities = [max(dists.values()) for (source, dists) in sorted(paths)]
local_measures = np.concatenate(
[[v for (k, v) in sorted(clustering_dict.items())],
[v for (k, v) in sorted(betweenness_centrality_dict.items())],
eccentricities])
graph_diameter = max(eccentricities)
graph_radius = min(eccentricities)
aspl = average_shortest_path_length(G, weight='weight')
global_measures = np.array([graph_diameter, graph_radius, aspl])
return np.concatenate([local_measures, global_measures])
def getNodeClusteringCoefficient(
self,
node: V) -> Union[Any, int, float, dict[Any, Union[Any, int]]]:
"""
"""
return clustering(self.graph, node)
G.add_edge(edge, node_id)
pG = projected_graph(G, people)
distances_to_furthest_nodes = dict()
for source, targets_to_paths in shortest_path(pG).iteritems():
longest_shortest_path_length = 0
for target, path in targets_to_paths.iteritems():
longest_shortest_path_length = max(
[longest_shortest_path_length,
len(path)])
distances_to_furthest_nodes[source] = longest_shortest_path_length
nodes['distanceToFurthestNode'] = pandas.Series(distances_to_furthest_nodes,
index=nodes.index)
nodes['clusteringCoefficient'] = pandas.Series(clustering(pG),
index=nodes.index)
nodes['betweennessCentrality'] = pandas.Series(betweenness_centrality(pG),
index=nodes.index)
nodes['degree'] = pandas.Series(pG.degree(), index=nodes.index)
print "vertices: {}".format(pG.number_of_nodes())
print "edges: {}".format(pG.number_of_edges())
print nodes.nlargest(
5, 'clusteringCoefficient')[['clusteringCoefficient', 'name']]
print nodes.nlargest(
5, 'betweennessCentrality')[['betweennessCentrality', 'name']]
print nodes.nlargest(
5, 'distanceToFurthestNode')[['distanceToFurthestNode', 'name']]
print nodes.nlargest(5, 'degree')[['degree', 'name']]
def ver_medidas(G):
print(function.info(G))
"""
Numero minimo de nodos que deben ser removidos para desconectar G
"""
print("Numero minimo de nodos que deben ser removidos para desconectar G :"+str(approximation.node_connectivity(G)))
"""
average clustering coefficient of G.
"""
print("average clustering coefficient of G: "+str(approximation.average_clustering(G)))
"""
Densidad de un Grafo
"""
print("Densidad de G: "+str(function.density(G)))
"""
Assortativity measures the similarity of connections in
the graph with respect to the node degree.
Valores positivos de r indican que existe una correlacion entre nodos
con grado similar, mientras que un valor negativo indica
correlaciones entre nodos de diferente grado
"""
print("degree assortativity:"+str(assortativity.degree_assortativity_coefficient(G)))
"""
Assortativity measures the similarity of connections
in the graph with respect to the given attribute.
"""
print("assortativity for node attributes: "+str(assortativity.attribute_assortativity_coefficient(G,"crime")))
"""
Grado promedio vecindad
"""
plt.plot(assortativity.average_neighbor_degree(G).values())
plt.title("Grado promedio vecindad")
plt.xlabel("Nodo")
plt.ylabel("Grado")
plt.show();
"""
Grado de Centralidad de cada nodo
"""
plt.plot(centrality.degree_centrality(G).values())
plt.title("Grado de centralidad")
plt.xlabel("Nodo")
plt.ylabel("Centralidad")
plt.show();
"""
Calcular el coeficiente de agrupamiento para nodos
"""
plt.plot(cluster.clustering(G).values())
plt.title("coeficiente de agrupamiento")
plt.xlabel("Nodo")
plt.show();
"""
Media coeficiente de Agrupamiento
"""
print("Coeficiente de agrupamiento de G:"+str(cluster.average_clustering(G)))
"""
Centro del grafo
El centro de un grafo G es el subgrafo inducido por el
conjunto de vertices de excentricidad minima.
La excentricidad de v in V se define como la
distancia maxima desde v a cualquier otro vertice del
grafo G siguiendo caminos de longitud minima.
"""
print("Centro de G:"+ str(distance_measures.center(G)))
"""
Diametro de un grafo
The diameter is the maximum eccentricity.
"""
print("Diametro de G:"+str(distance_measures.diameter(G)))
"""
Excentricidad de cada Nodo
The eccentricity of a node v is the maximum distance
from v to all other nodes in G.
"""
plt.plot(distance_measures.eccentricity(G).values())
plt.title("Excentricidad de cada Nodo")
plt.xlabel("Nodo")
plt.show();
"""
Periferia
The periphery is the set of nodes with eccentricity equal to the diameter.
"""
print("Periferia de G:")
print(distance_measures.periphery(G))
"""
Radio
The radius is the minimum eccentricity.
"""
print("Radio de G:"+str(distance_measures.radius(G)))
"""
PageRank calcula una clasificacion de los nodos
en el grafico G en funcion de la estructura de
los enlaces entrantes. Originalmente fue disenado
como un algoritmo para clasificar paginas web.
"""
plt.plot(link_analysis.pagerank_alg.pagerank(G).values())
plt.title("Puntaje de cada Nodo")
plt.xlabel("Nodo")
plt.show();
"""
Coeficiente de Small World.
A graph is commonly classified as small-world if sigma>1.
"""
print("Coeficiente de Small World: " + str(smallworld.sigma(G)))
"""
The small-world coefficient (omega) ranges between -1 and 1.
Values close to 0 means the G features small-world characteristics.
Values close to -1 means G has a lattice shape whereas values close
to 1 means G is a random graph.
"""
print("Omega coeficiente: "+str(smallworld.omega(G)))
y = np.array(y) + 1
x = np.log(x)
y = np.log(y)
plt.scatter(x, y, s=1, color=(1, 0, 0))
plt.show()
net_file = '../net_6'
# Read graph from file
G = read_edgelist(net_file)
print("The net_6 has %d nodes and %d edges" %
(G.number_of_nodes(), G.number_of_edges()))
# Clustering coefficient 0.07872
C = list(clustering(G).values())
C = sum(C) / len(C)
print("The net_6 has clustering coefficient: %f" % C)
scatterplot_degree_distribution(G)
log_scatterplot_degree_distribution(G)
d = parallel_diam(G, 5)
print("The net_6 has diameter: ", d)
###############################
## net 6 is random?
###############################
RG = randomG(10000, 0.0016)
print("The random graph has %d nodes and %d edges" %
(G.number_of_nodes(), G.number_of_edges()))
# Clustering coefficient 0.07872
# deg_dict = OrderedDict(sorted(graph.degree(), key=lambda x: x[0]))
# deg_dict2 = OrderedDict(sorted(graph2.degree(), key=lambda x: x[0]))
# deg_dict3 = OrderedDict(sorted(graph3.degree(), key=lambda x: x[0]))
# deg_dict_avg = [(list(deg_dict.values())[x] + list(deg_dict2.values())[x] + list(deg_dict3.values())[x])/3 for x in range(len(deg_dict.values()))]
#
# # Assortativity Plot
# plt.scatter(deg_dict_avg, neighbor_deg_avg, marker=".")
# plt.xlabel('Degree')
# plt.ylabel('Average neighbour degree')
# plt.show()
# # Pearson Correlation of Plot:
# print(numpy.corrcoef(list(deg_dict_avg), list(neighbor_deg_avg))[0, 1])
# Clustering Data
data = OrderedDict(sorted((cluster.clustering(graph)).items()))
data2 = OrderedDict(sorted((cluster.clustering(graph2)).items()))
data3 = OrderedDict(sorted((cluster.clustering(graph3)).items()))
# Clustering Calculation
values = [(list(data.values())[x] + list(data2.values())[x] + list(data3.values())[x])/3 for x in range(len(data.values()))]
values.sort()
keys = data.keys()
# Clustering Plot
plt.scatter(values, range(len(keys)), marker=".")
plt.ylabel('Cumulative frequency of Nodes')
plt.xlabel('Local Clustering Coefficient Value of Nodes')
# Twitter Annotation:
plt.annotate('Median value at 0.4 and 500', xytext=(0.45, 500), xy=(0.402, 500), arrowprops = {'facecolor':'red'})
# # Google+ Annotation:
def avgClustering(self, W):
from networkx.algorithms.cluster import clustering
g = buildGraph(W, weighted=self.weighted)
cl = clustering(g, weight='weight' if self.weighted else None)
return centerMeasure(cl.values(), self.center)
def find_topics_topicness(g: nx.DiGraph, visualize=False) -> np.ndarray:
"""
Computes a keyword-topic matrix containing memebership information of keywords w.r.t topics.
Extended SUmmuray
-----------------
The computation of this matrix uses an iterative algorithm, which picks best nodes using a 'topicness'
metric, derived from pagerank, local cluster coefficient, and betweenness centrality.
Parameters
----------
g : nx.DiGraph
keyword co-occurrence graph
Returns
-------
ndarray
keyword-topic matrix, containing the information about keywords and topics.
specifically for the entry [ui,ti] if contains 1 if the topic is in
"""
n = g.number_of_nodes()
m = g.number_of_edges()
# get betweenness centrality
betweenness = misc.dictionary_to_numpy(
nx.betweenness_centrality(g.to_undirected(as_view=True)))
# get personalized-pagerank using cluster coefficient
c = cluster.clustering(g, weight="weight")
z = sum(c.values())
if z != 0:
v = {ui: c[ui] / z for ui in range(n)}
pagerank = misc.dictionary_to_numpy(
nx.pagerank_numpy(g, personalization=v))
else:
pagerank = misc.dictionary_to_numpy(nx.pagerank_numpy(g))
# compute topicness --------------------------
pagerank: np.ndarray = pagerank / pagerank.max() # normalize between [0,1]
betweenness = betweenness / betweenness.max() if betweenness.max(
) != 0 else betweenness # normalize between [0,1]
topicness: np.ndarray = pagerank * (1 - betweenness)
# normalize topicness between [0,1]
topicness = topicness - topicness.min()
topicness = topicness / (topicness.max())
# normalize the edge weights
influence.normalize_weights(g, mode="mixed")
# compute area of influence for each node
node_influence = np.zeros([n, n], dtype=float)
for u in tqdm.trange(n):
node_influence[u, :] = influence.linear_threshold_mean(g, {u}, 15)
# estiamte topics
topics_list, surface = list(), np.array(topicness)
#for i in tqdm.trange(k):
while (surface > 0).any():
# compute topic influence
fuzzy_topic = node_influence[surface.argmax(), :]
crisp_topic = defuzzification(fuzzy_topic)
# add topic to extracted topics
topics_list.append(crisp_topic)
# update surface
surface = surface * (1 - fuzzy_topic)
# fill node-topic matrix
topic_number = len(topics_list)
T = np.zeros([n, topic_number], dtype=float) # node-topic matrix
for i in range(topic_number):
T[:, i] = topics_list[i]
# visualize various information about the estiamted topics
if visualize:
coverage = T.sum(axis=1)
sources = np.zeros([n])
sources[T.argmax(axis=0)] = 1
visualization.plot_edge_weights(g)
visualization.show_graphfunction(g, topicness, with_labels=False)
visualization.show_graphfunction(g, sources, with_labels=False)
visualization.show_graphfunction(g, coverage, with_labels=False)
return T
# coding: utf-8 import networkx import igraph graph_nx = networkx.waxman_graph(20) graph_ig = igraph.Graph(len(graph_nx), list(zip(*list(zip(*networkx.to_edgelist(graph_nx)))[:2]))) import numpy as np import networkx.algorithms.cluster as nx_cluster cluster_nx = np.mean(list(nx_cluster.clustering(graph_nx).values())) cluster_nx nx_cluster.clustering(graph_nx).values() graph_nx = networkx.waxman_graph(100) nx_cluster.clustering(graph_nx).values() cluster_nx = np.mean(list(nx_cluster.clustering(graph_nx).values())) cluster_nx graph_ig = igraph.Graph(len(graph_nx), list(zip(*list(zip(*networkx.to_edgelist(graph_nx)))[:2]))) graph_ig.transitivity_avglocal_undirected() graph_ig.transitivity_undirected() graph_ig.transitivity_local_undirected() cluster_ig = np.mean(graph_ig.transitivity_local_undirected()) cluster_ig np.mean(graph_ig.transitivity_local_undirected()) graph_ig.transitivity_local_undirected() cluster_ig = graph_ig.transitivity_local_undirected() cluster_ig np.nan_to_num(cluster_ig) cluster_ig cluster_ig = np.nan_to_num(cluster_ig) cluster_ig np.mean(cluster_ig) nx_cluster.transitivity(graph_nx)
