def test_edmot():
"""
Test the EdMot procedure.
"""
graph = nx.newman_watts_strogatz_graph(50, 5, 0.3)
model = EdMot()
model.fit(graph)
memberships = model.get_memberships()
indices = [k for k, v in memberships.items()].sort()
nodes = [node for node in graph.nodes()].sort()
assert graph.number_of_nodes() == len(memberships)
assert indices == nodes
assert type(memberships) == dict
graph = nx.newman_watts_strogatz_graph(150, 5, 0.3)
model = EdMot()
model.fit(graph)
memberships = model.get_memberships()
indices = [k for k, v in memberships.items()].sort()
nodes = [node for node in graph.nodes()].sort()
assert graph.number_of_nodes() == len(memberships)
assert indices == nodes
assert type(memberships) == dict
for num in nums:
reader = GraphReader("twitch%s" % num)
graph = reader.get_graph()
target = reader.get_target()
d = {}
model = LabelPropagation()
comms = run(model, graph)
coeffs = []
for key in sorted(comms.keys()):
sub = graph.subgraph(list(comms[key]))
coeff = nx.average_clustering(sub)
coeffs.append(coeff)
avg_coeff = sum(coeffs) / len(coeffs)
d['label'] = avg_coeff
model = EdMot()
comms = run(model, graph)
coeffs = []
for key in sorted(comms.keys()):
sub = graph.subgraph(list(comms[key]))
coeff = nx.average_clustering(sub)
coeffs.append(coeff)
avg_coeff = sum(coeffs) / len(coeffs)
d['edmot'] = avg_coeff
model = GEMSEC()
comms = run(model, graph)
coeffs = []
for key in sorted(comms.keys()):
sub = graph.subgraph(list(comms[key]))
coeff = nx.average_clustering(sub)
# Splitter example #------------------- g = nx.newman_watts_strogatz_graph(100, 20, 0.05) model = EgoNetSplitter(1.0) model.fit(g) #---------------- # Edmot example #---------------- g = nx.newman_watts_strogatz_graph(100, 10, 0.9) model = EdMot(3, 0.5) model.fit(g) #---------------- # DANMF example #---------------- g = nx.newman_watts_strogatz_graph(100, 10, 0.02) model = DANMF() model.fit(g) #---------------- # M-NMF example
"""EdMot example.""" import networkx as nx from karateclub.community_detection.non_overlapping import EdMot g = nx.newman_watts_strogatz_graph(100, 20, 0.05) model = EdMot() model.fit(g)
def edmot(population, neighbors, probability):
g = nx.newman_watts_strogatz_graph(population, neighbors, probability)
model = EdMot(3, 0.5)
model.fit(g)
#print(model.get_memberships())
return [model.get_memberships()]
from karateclub.node_embedding.neighbourhood import GraRep, DeepWalk, Walklets, NMFADMM, Diff2Vec, BoostNE, NetMF
from karateclub.community_detection.overlapping import EgoNetSplitter, NNSED, DANMF, MNMF, BigClam, SymmNMF
from karateclub.community_detection.non_overlapping import EdMot, LabelPropagation, SCD
from karateclub.graph_embedding import Graph2Vec, FGSD, GL2Vec, SF
from karateclub.node_embedding.attributed import BANE, TENE, TADW, FSCNMF, SINE, MUSAE
from karateclub.node_embedding.structural import GraphWave, Role2Vec
from karateclub.dataset import GraphReader, GraphSetReader
#------------------------------------
# Edmot example
#------------------------------------
g = nx.newman_watts_strogatz_graph(100, 10, 0.1)
model = EdMot(3, 4)
model.fit(g)
#------------------------------------
# MUSAE example
#------------------------------------
g = nx.newman_watts_strogatz_graph(100, 10, 0.2)
X = {i: random.sample(range(150), 50) for i in range(100)}
row = np.array([k for k, v in X.items() for val in v])
col = np.array([val for k, v in X.items() for val in v])
data = np.ones(100 * 50)
shape = (100, 150)