def test_egonet_splitter():
"""
Test the Ego Net splitter procedure.
"""
graph = nx.newman_watts_strogatz_graph(100, 5, 0.3)
model = EgoNetSplitter()
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 = EgoNetSplitter()
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
# Test weighted graph
graph = nx.les_miserables_graph()
graph = nx.convert_node_labels_to_integers(graph)
model = EgoNetSplitter(weight='weight')
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
# Force unweighted
graph = nx.les_miserables_graph()
graph = nx.convert_node_labels_to_integers(graph)
model = EgoNetSplitter(weight=None)
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
def prepare_data(
) -> Tuple[PropagationModel, Dict[str, Tuple[str, ...]], MultilayerNetwork]:
"""Sets up data needed to perform tests."""
# init propagation model and set transitions with probabilities
model = PropagationModel()
phenomenas = {
"ill": ("S", "I", "R"),
"aware": ("UA", "A"),
"vacc": ("UV", "V"),
}
for l, p in phenomenas.items():
model.add(l, p)
model.compile(background_weight=0)
# init multilayer network from nx predefined network
network = MultilayerNetwork()
network.load_layer_nx(nx.les_miserables_graph(), [*phenomenas.keys()])
model.set_transition_fast("ill.S", "ill.I", ("vacc.UV", "aware.UA"), 0.9)
model.set_transition_fast("ill.S", "ill.I", ("vacc.V", "aware.A"), 0.05)
model.set_transition_fast("ill.S", "ill.I", ("vacc.UV", "aware.A"), 0.2)
model.set_transition_fast("ill.I", "ill.R", ("vacc.UV", "aware.UA"), 0.1)
model.set_transition_fast("ill.I", "ill.R", ("vacc.V", "aware.A"), 0.7)
model.set_transition_fast("ill.I", "ill.R", ("vacc.UV", "aware.A"), 0.3)
model.set_transition_fast("vacc.UV", "vacc.V", ("aware.A", "ill.S"), 0.03)
model.set_transition_fast("vacc.UV", "vacc.V", ("aware.A", "ill.I"), 0.01)
model.set_transition_fast("aware.UA", "aware.A", ("vacc.UV", "ill.S"),
0.05)
model.set_transition_fast("aware.UA", "aware.A", ("vacc.V", "ill.S"), 1)
model.set_transition_fast("aware.UA", "aware.A", ("vacc.UV", "ill.I"), 0.2)
return model, phenomenas, network
def setup_class(cls):
G = nx.Graph()
G.add_edge(0, 1, weight=3)
G.add_edge(0, 2, weight=2)
G.add_edge(0, 3, weight=6)
G.add_edge(0, 4, weight=4)
G.add_edge(1, 3, weight=5)
G.add_edge(1, 5, weight=5)
G.add_edge(2, 4, weight=1)
G.add_edge(3, 4, weight=2)
G.add_edge(3, 5, weight=1)
G.add_edge(4, 5, weight=4)
cls.G = G
cls.exact_weighted = {0: 4.0, 1: 0.0, 2: 8.0, 3: 6.0, 4: 8.0, 5: 0.0}
cls.K = nx.krackhardt_kite_graph()
cls.P3 = nx.path_graph(3)
cls.P4 = nx.path_graph(4)
cls.K5 = nx.complete_graph(5)
cls.C4 = nx.cycle_graph(4)
cls.T = nx.balanced_tree(r=2, h=2)
cls.Gb = nx.Graph()
cls.Gb.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3), (2, 4), (4, 5),
(3, 5)])
cls.F = nx.florentine_families_graph()
cls.LM = nx.les_miserables_graph()
cls.D = nx.cycle_graph(3, create_using=nx.DiGraph())
cls.D.add_edges_from([(3, 0), (4, 3)])
def setUp(self):
G = nx.Graph()
G.add_edge(0, 1, weight=3)
G.add_edge(0, 2, weight=2)
G.add_edge(0, 3, weight=6)
G.add_edge(0, 4, weight=4)
G.add_edge(1, 3, weight=5)
G.add_edge(1, 5, weight=5)
G.add_edge(2, 4, weight=1)
G.add_edge(3, 4, weight=2)
G.add_edge(3, 5, weight=1)
G.add_edge(4, 5, weight=4)
self.G = G
self.exact_weighted = {0: 4.0, 1: 0.0, 2: 8.0, 3: 6.0, 4: 8.0, 5: 0.0}
self.K = nx.krackhardt_kite_graph()
self.P3 = nx.path_graph(3)
self.P4 = nx.path_graph(4)
self.K5 = nx.complete_graph(5)
self.C4 = nx.cycle_graph(4)
self.T = nx.balanced_tree(r=2, h=2)
self.Gb = nx.Graph()
self.Gb.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3),
(2, 4), (4, 5), (3, 5)])
self.F = nx.florentine_families_graph()
self.LM = nx.les_miserables_graph()
self.D = nx.cycle_graph(3, create_using=nx.DiGraph())
self.D.add_edges_from([(3, 0), (4, 3)])
def test_les_miserables_graph_cutset(self):
G = nx.les_miserables_graph()
nx.set_edge_attributes(G, 1, "capacity")
for flow_func in flow_funcs:
T = nx.gomory_hu_tree(G, flow_func=flow_func)
assert nx.is_tree(T)
for u, v in combinations(G, 2):
cut_value, edge = self.minimum_edge_weight(T, u, v)
assert nx.minimum_cut_value(G, u, v) == cut_value
def test_les_miserables_graph_cutset(self):
G = nx.les_miserables_graph()
nx.set_edge_attributes(G, 1, 'capacity')
for flow_func in flow_funcs:
T = nx.gomory_hu_tree(G, flow_func=flow_func)
assert_true(nx.is_tree(T))
for u, v in combinations(G, 2):
cut_value, edge = self.minimum_edge_weight(T, u, v)
assert_equal(nx.minimum_cut_value(G, u, v),
cut_value)
def setUp(self):
self.K = nx.krackhardt_kite_graph()
self.P3 = nx.path_graph(3)
self.P4 = nx.path_graph(4)
self.K5 = nx.complete_graph(5)
self.C4 = nx.cycle_graph(4)
self.T = nx.balanced_tree(r=2, h=2)
self.Gb = nx.Graph()
self.Gb.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3), (2, 4), (4, 5),
(3, 5)])
F = nx.florentine_families_graph()
self.F = F
self.LM = nx.les_miserables_graph()
def setUp(self):
self.K = nx.krackhardt_kite_graph()
self.P3 = nx.path_graph(3)
self.P4 = nx.path_graph(4)
self.K5 = nx.complete_graph(5)
self.C4 = nx.cycle_graph(4)
self.T = nx.balanced_tree(r=2, h=2)
self.Gb = nx.Graph()
self.Gb.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3),
(2, 4), (4, 5), (3, 5)])
F = nx.florentine_families_graph()
self.F = F
self.LM = nx.les_miserables_graph()
def setUp(self) -> None:
"""Sets up testing parameters and performs simulation."""
self.exp_M = nx.les_miserables_graph()
self.exp_beta = 0.1
self.exp_I = 0.05
self.exp_name = "".join(random.choices(string.ascii_letters, k=5))
self.out_dir = "".join(random.choices(string.ascii_letters, k=5))
(
self.list_S,
self.list_I,
self.list_iter,
self.nodes_infected,
self.par,
) = FlatSpreading.si_diffusion(self.exp_M,
self.exp_I,
self.exp_beta,
name=self.exp_name)
def setup_class(cls):
cls.K = nx.krackhardt_kite_graph()
cls.P3 = nx.path_graph(3)
cls.P4 = nx.path_graph(4)
cls.K5 = nx.complete_graph(5)
cls.C4 = nx.cycle_graph(4)
cls.T = nx.balanced_tree(r=2, h=2)
cls.Gb = nx.Graph()
cls.Gb.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3), (2, 4), (4, 5), (3, 5)])
F = nx.florentine_families_graph()
cls.F = F
cls.LM = nx.les_miserables_graph()
# Create random undirected, unweighted graph for testing incremental version
cls.undirected_G = nx.fast_gnp_random_graph(n=100, p=0.6, seed=123)
cls.undirected_G_cc = nx.closeness_centrality(cls.undirected_G)
def test_compute_multiplexing_coefficient(self) -> None:
"""Tests if multiplexing coefficient is computed correctly."""
network = MultilayerNetwork()
names = ["a", "b"]
network.load_layer_nx(nx.les_miserables_graph(), names)
mcf = network._compute_multiplexing_coefficient()
self.assertEqual(
mcf,
1,
"Multiplexing coefficient for multpilex network should be 1",
)
mcf = self.network._compute_multiplexing_coefficient()
self.assertAlmostEqual(
mcf,
0.733,
2,
'Multiplexing coefficient for "florentine" network should be '
"almost 0.73",
)
def test_load_layer_nx(self) -> None:
"""Checks if creating MLN is correct by load_layer_nx func."""
network = MultilayerNetwork()
names = ["a", "b"]
network.load_layer_nx(nx.les_miserables_graph(), names)
self.assertTrue(
network.layers["a"].nodes == network.layers["b"].nodes,
'Nodes of layer "a" and "b" should be equal!',
)
self.assertTrue(
network.layers["a"].edges == network.layers["b"].edges,
'Edges of layer "a" and "b" should be equal!',
)
self.assertFalse(
network.layers["a"] == network.layers["b"],
'Network in layer "a" should be a diffeereent object than in "b"!',
)
self.assertEqual(
network.layers["a"].nodes["Napoleon"]["status"],
None,
"Node should have None in status attr",
)
import networkx as nx
from network_diffusion.multilayer_network import MultilayerNetwork
from network_diffusion.propagation_model import PropagationModel
from network_diffusion.multi_spreading import MultiSpreading
from os import getcwd
# first example
# initialise multilayer network from nx prefedined network
network = MultilayerNetwork()
names = ['illness', 'awareness', 'vaccination']
network.load_layer_nx(nx.les_miserables_graph(), names)
# initialise propagation model and set possible transitions with probabilities
model = PropagationModel()
phenomenas = [('S', 'I', 'R'), ('UA', 'A'), ('UV', 'V')]
for l, p in zip(names, phenomenas):
model.add(l, p)
model.compile(background_weight=0.005)
model.set_transition('illness.S', 'illness.I',
['vaccination.UV', 'awareness.UA'], 0.9)
model.set_transition('illness.S', 'illness.I',
['vaccination.V', 'awareness.A'], 0.05)
model.set_transition('illness.S', 'illness.I',
['vaccination.UV', 'awareness.A'], 0.2)
model.set_transition('illness.I', 'illness.R',
['vaccination.UV', 'awareness.UA'], 0.1)
model.set_transition('illness.I', 'illness.R',
['vaccination.V', 'awareness.A'], 0.7)
model.set_transition('illness.I', 'illness.R',
fig, ax = plt.subplots(3, 4, figsize=(15,12))
G = nx.scale_free_graph(50)
nx.draw(nx.complete_graph(50), ax=ax[0,0])
nx.draw(nx.star_graph(50), ax=ax[0,1])
nx.draw(nx.circular_ladder_graph(50), ax=ax[0,2])
nx.draw(nx.ladder_graph(50), ax=ax[0,3])
nx.draw(nx.path_graph(50), ax=ax[1,0])
nx.draw(nx.wheel_graph(50), ax=ax[1,1])
nx.draw(nx.erdos_renyi_graph(50, 0.3), ax=ax[1,2])
nx.draw(nx.barabasi_albert_graph(50, 5), ax=ax[1,3])
nx.draw(nx.random_powerlaw_tree(10), ax=ax[2,0])
nx.draw(nx.scale_free_graph(50), ax=ax[2,1])
nx.draw(nx.karate_club_graph(), ax=ax[2,2])
nx.draw(nx.les_miserables_graph(), ax=ax[2,3])
# %% [markdown]
# ## Network Statistics
# Calculate basic network statistics
# %%
# import an edgelist as a weighted directed graph
filepath = r"data/sample.edgelist"
G = nx.read_edgelist(filepath, create_using=nx.DiGraph,data=(('weight', int),), delimiter=' ')
# %% [markdown]
# This is a fairly large network
# %%
def les_miserables():
G = nx.les_miserables_graph()
print("les_miserables_graph => |V| =", G.number_of_nodes())
return G
def test_les_miserables_graph(self):
"""Weighted betweenness centrality: Les Miserables graph"""
G = nx.les_miserables_graph()
b_answer = {
"Napoleon": 0.000,
"Myriel": 0.177,
"MlleBaptistine": 0.000,
"MmeMagloire": 0.000,
"CountessDeLo": 0.000,
"Geborand": 0.000,
"Champtercier": 0.000,
"Cravatte": 0.000,
"Count": 0.000,
"OldMan": 0.000,
"Valjean": 0.454,
"Labarre": 0.000,
"Marguerite": 0.009,
"MmeDeR": 0.000,
"Isabeau": 0.000,
"Gervais": 0.000,
"Listolier": 0.000,
"Tholomyes": 0.066,
"Fameuil": 0.000,
"Blacheville": 0.000,
"Favourite": 0.000,
"Dahlia": 0.000,
"Zephine": 0.000,
"Fantine": 0.114,
"MmeThenardier": 0.046,
"Thenardier": 0.129,
"Cosette": 0.075,
"Javert": 0.193,
"Fauchelevent": 0.026,
"Bamatabois": 0.080,
"Perpetue": 0.000,
"Simplice": 0.001,
"Scaufflaire": 0.000,
"Woman1": 0.000,
"Judge": 0.000,
"Champmathieu": 0.000,
"Brevet": 0.000,
"Chenildieu": 0.000,
"Cochepaille": 0.000,
"Pontmercy": 0.023,
"Boulatruelle": 0.000,
"Eponine": 0.023,
"Anzelma": 0.000,
"Woman2": 0.000,
"MotherInnocent": 0.000,
"Gribier": 0.000,
"MmeBurgon": 0.026,
"Jondrette": 0.000,
"Gavroche": 0.285,
"Gillenormand": 0.024,
"Magnon": 0.005,
"MlleGillenormand": 0.036,
"MmePontmercy": 0.005,
"MlleVaubois": 0.000,
"LtGillenormand": 0.015,
"Marius": 0.072,
"BaronessT": 0.004,
"Mabeuf": 0.089,
"Enjolras": 0.003,
"Combeferre": 0.000,
"Prouvaire": 0.000,
"Feuilly": 0.004,
"Courfeyrac": 0.001,
"Bahorel": 0.007,
"Bossuet": 0.028,
"Joly": 0.000,
"Grantaire": 0.036,
"MotherPlutarch": 0.000,
"Gueulemer": 0.025,
"Babet": 0.015,
"Claquesous": 0.042,
"Montparnasse": 0.050,
"Toussaint": 0.011,
"Child1": 0.000,
"Child2": 0.000,
"Brujon": 0.002,
"MmeHucheloup": 0.034,
}
b = nx.betweenness_centrality(G, weight="weight", normalized=True)
for n in sorted(G):
assert almost_equal(b[n], b_answer[n], places=3)
import matplotlib.pyplot as plt
import networkx as nx
from hc_embedding import draw_hce
if __name__ == '__main__':
graphs = [
nx.karate_club_graph(),
nx.davis_southern_women_graph(),
nx.les_miserables_graph(),
nx.florentine_families_graph(),
nx.powerlaw_cluster_graph(250, 1, 0.001),
]
labels = [
"Karate",
"Davis Southern Women",
"Les Miserables",
"Florentine Families",
"Power Law Cluster",
]
for G, title in zip(graphs, labels):
draw_hce(G, title)
plt.savefig(f"{title.lower().replace(' ', '_')}_test.png")
plt.close()
def test_les_miserables_graph(self):
"""Weighted betweenness centrality: Les Miserables graph"""
G = nx.les_miserables_graph()
b_answer = \
{'Napoleon': 0.000,
'Myriel': 0.177,
'MlleBaptistine': 0.000,
'MmeMagloire': 0.000,
'CountessDeLo': 0.000,
'Geborand': 0.000,
'Champtercier': 0.000,
'Cravatte': 0.000,
'Count': 0.000,
'OldMan': 0.000,
'Valjean': 0.454,
'Labarre': 0.000,
'Marguerite': 0.009,
'MmeDeR': 0.000,
'Isabeau': 0.000,
'Gervais': 0.000,
'Listolier': 0.000,
'Tholomyes': 0.066,
'Fameuil': 0.000,
'Blacheville': 0.000,
'Favourite': 0.000,
'Dahlia': 0.000,
'Zephine': 0.000,
'Fantine': 0.114,
'MmeThenardier': 0.046,
'Thenardier': 0.129,
'Cosette': 0.075,
'Javert': 0.193,
'Fauchelevent': 0.026,
'Bamatabois': 0.080,
'Perpetue': 0.000,
'Simplice': 0.001,
'Scaufflaire': 0.000,
'Woman1': 0.000,
'Judge': 0.000,
'Champmathieu': 0.000,
'Brevet': 0.000,
'Chenildieu': 0.000,
'Cochepaille': 0.000,
'Pontmercy': 0.023,
'Boulatruelle': 0.000,
'Eponine': 0.023,
'Anzelma': 0.000,
'Woman2': 0.000,
'MotherInnocent': 0.000,
'Gribier': 0.000,
'MmeBurgon': 0.026,
'Jondrette': 0.000,
'Gavroche': 0.285,
'Gillenormand': 0.024,
'Magnon': 0.005,
'MlleGillenormand': 0.036,
'MmePontmercy': 0.005,
'MlleVaubois': 0.000,
'LtGillenormand': 0.015,
'Marius': 0.072,
'BaronessT': 0.004,
'Mabeuf': 0.089,
'Enjolras': 0.003,
'Combeferre': 0.000,
'Prouvaire': 0.000,
'Feuilly': 0.004,
'Courfeyrac': 0.001,
'Bahorel': 0.007,
'Bossuet': 0.028,
'Joly': 0.000,
'Grantaire': 0.036,
'MotherPlutarch': 0.000,
'Gueulemer': 0.025,
'Babet': 0.015,
'Claquesous': 0.042,
'Montparnasse': 0.050,
'Toussaint': 0.011,
'Child1': 0.000,
'Child2': 0.000,
'Brujon': 0.002,
'MmeHucheloup': 0.034}
b = nx.betweenness_centrality(G, weight='weight', normalized=True)
for n in sorted(G):
assert_almost_equal(b[n], b_answer[n], places=3)
def test_les_miserables_graph(self):
"""Weighted betweenness centrality: Les Miserables graph"""
G = nx.les_miserables_graph()
b_answer = \
{'Napoleon': 0.000,
'Myriel': 0.177,
'MlleBaptistine': 0.000,
'MmeMagloire': 0.000,
'CountessDeLo': 0.000,
'Geborand': 0.000,
'Champtercier': 0.000,
'Cravatte': 0.000,
'Count': 0.000,
'OldMan': 0.000,
'Valjean': 0.454,
'Labarre': 0.000,
'Marguerite': 0.009,
'MmeDeR': 0.000,
'Isabeau': 0.000,
'Gervais': 0.000,
'Listolier': 0.000,
'Tholomyes': 0.066,
'Fameuil': 0.000,
'Blacheville': 0.000,
'Favourite': 0.000,
'Dahlia': 0.000,
'Zephine': 0.000,
'Fantine': 0.114,
'MmeThenardier': 0.046,
'Thenardier': 0.129,
'Cosette': 0.075,
'Javert': 0.193,
'Fauchelevent': 0.026,
'Bamatabois': 0.080,
'Perpetue': 0.000,
'Simplice': 0.001,
'Scaufflaire': 0.000,
'Woman1': 0.000,
'Judge': 0.000,
'Champmathieu': 0.000,
'Brevet': 0.000,
'Chenildieu': 0.000,
'Cochepaille': 0.000,
'Pontmercy': 0.023,
'Boulatruelle': 0.000,
'Eponine': 0.023,
'Anzelma': 0.000,
'Woman2': 0.000,
'MotherInnocent': 0.000,
'Gribier': 0.000,
'MmeBurgon': 0.026,
'Jondrette': 0.000,
'Gavroche': 0.285,
'Gillenormand': 0.024,
'Magnon': 0.005,
'MlleGillenormand': 0.036,
'MmePontmercy': 0.005,
'MlleVaubois': 0.000,
'LtGillenormand': 0.015,
'Marius': 0.072,
'BaronessT': 0.004,
'Mabeuf': 0.089,
'Enjolras': 0.003,
'Combeferre': 0.000,
'Prouvaire': 0.000,
'Feuilly': 0.004,
'Courfeyrac': 0.001,
'Bahorel': 0.007,
'Bossuet': 0.028,
'Joly': 0.000,
'Grantaire': 0.036,
'MotherPlutarch': 0.000,
'Gueulemer': 0.025,
'Babet': 0.015,
'Claquesous': 0.042,
'Montparnasse': 0.050,
'Toussaint': 0.011,
'Child1': 0.000,
'Child2': 0.000,
'Brujon': 0.002,
'MmeHucheloup': 0.034}
b = nx.betweenness_centrality(G,
weight='weight',
normalized=True)
for n in sorted(G):
assert_almost_equal(b[n], b_answer[n], places=3)
