def test_only_critical(self):
# tests only critical vertices in form of trees, paths and stars,
# expected the algorithm correctly augments G and the augmenting set
# cardinality correspond the simple bound on Eswaran-Tarjan.
D: nx.DiGraph = nx.DiGraph()
nx.add_star(D, {i for i in range(1, 5)})
nx.add_path(D, {i for i in range(5, 10)})
D.add_nodes_from({i for i in range(10, 20)})
G, A, M = D_to_bipartite(D)
assert_true(is_correctly_augmented(G, A))
sources, sinks, isolated = get_sources_sinks_isolated(D)
s, t, q = len(sources), len(sinks), len(isolated)
L = bipartite_matching_augmentation(G, A)
assert_equal(len(L), max(s, t) + q)
D.clear()
D = nx.balanced_tree(2, 13, nx.DiGraph())
D.remove_node(0)
G, A, M = D_to_bipartite(D)
sources, sinks, isolated = get_sources_sinks_isolated(D)
s, t, q = len(sources), len(sinks), len(isolated)
for i in range(1):
L = bipartite_matching_augmentation(G, A)
assert_true(is_correctly_augmented(G, A, L))
assert_equal(len(L), max(s, t) + q)
def topology():
"""
Function used for create new Graph.
Execute this func first.
"""
try:
attr = param()
core_nodes = attr['core_nodes']
total_ce = attr['total_ce']
total_rr = attr['total_rr']
sum_nodes = attr['sum_nodes']
ls = attr['ls']
if total_ce > 0:
nodes_ce = sorted(
map(lambda x: 'CE{}'.format(x), range(0, total_ce)))
#Generate nodes with name 'RX', where X is serial number
nodes = sorted(map(lambda x: 'R{}'.format(x), range(0, core_nodes)))
# Peers for CE nodes
pe_peers = random.sample(nodes, total_ce)
if total_rr > 0:
pe_peers = random.sample(nodes[1:], total_ce)
G = nx.DiGraph(name='Network topology - Star')
nx.add_star(G, nodes)
gen_edge(G, nodes, total_rr, core_nodes, pe_peers, sum_nodes,
total_ce, nodes_ce, ls)
else:
G = nx.complete_graph(core_nodes)
gen_edge(G, nodes, total_rr, core_nodes, pe_peers, sum_nodes,
total_ce, nodes_ce, ls)
return G
except AddressValueError as value_error:
print('Error: {}'.format(value_error))
except ZeroDivisionError:
print('Not valid core_nodes. Allowed maximum 16 nodes!')
def setUp(self):
self.partitions = []
graph = nx.Graph()
nx.add_star(graph, range(20))
nx.add_path(graph, range(10))
self.partitions.append(
NxPartitionGraphBased(
graph, representation={node: node % 6
for node in graph}))
graph = nx.DiGraph()
nx.add_star(graph, range(20))
nx.add_path(graph, range(10))
self.partitions.append(
NxPartitionGraphBased(
graph, representation={node: node % 6
for node in graph}))
self.uniform_icl_ex = IntegratedCompleteLikelihoodExact(
is_directed=False,
hyperprior=IntegratedCompleteLikelihoodExact.UNIFORM_HYPERPRIOR)
self.jeffrey_icl_ex = IntegratedCompleteLikelihoodExact(
is_directed=False,
hyperprior=IntegratedCompleteLikelihoodExact.JEFFREY_HYPERPRIOR)
def test_star(self):
"""Closeness centrality: star """
G = nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b = nx.current_flow_closeness_centrality(G)
b_answer = {'a': 1.0 / 3, 'b': 0.6 / 3, 'c': 0.6 / 3, 'd': 0.6 / 3}
for n in sorted(G):
assert almost_equal(b[n], b_answer[n])
def test_star(self):
"""Betweenness centrality: star """
G = nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b = nx.current_flow_betweenness_centrality(G, normalized=True)
b_answer = {'a': 1.0, 'b': 0.0, 'c': 0.0, 'd': 0.0}
for n in sorted(G):
assert_almost_equal(b[n], b_answer[n])
def test_star(self):
"""Closeness centrality: star"""
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_closeness_centrality(G)
b_answer = {"a": 1.0 / 3, "b": 0.6 / 3, "c": 0.6 / 3, "d": 0.6 / 3}
for n in sorted(G):
assert b[n] == pytest.approx(b_answer[n], abs=1e-7)
def test_star(self):
"""Betweenness centrality: star """
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_betweenness_centrality_subset(G, list(G), list(G), normalized=True)
b_answer = nx.current_flow_betweenness_centrality(G, normalized=True)
for n in sorted(G):
assert_almost_equal(b[n], b_answer[n])
def test_star(self):
"""Closeness centrality: star """
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_closeness_centrality(G)
b_answer = {"a": 1.0 / 3, "b": 0.6 / 3, "c": 0.6 / 3, "d": 0.6 / 3}
for n in sorted(G):
assert almost_equal(b[n], b_answer[n])
def test_star(self):
"""Betweenness centrality: star"""
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_betweenness_centrality(G, normalized=True)
b_answer = {"a": 1.0, "b": 0.0, "c": 0.0, "d": 0.0}
for n in sorted(G):
assert almost_equal(b[n], b_answer[n])
def test_star(self):
"""Closeness centrality: star """
G=nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b=nx.current_flow_closeness_centrality(G)
b_answer={'a': 1.0/3, 'b': 0.6/3, 'c': 0.6/3, 'd':0.6/3}
for n in sorted(G):
assert_almost_equal(b[n],b_answer[n])
def test_star(self):
"""Betweenness centrality: star"""
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_betweenness_centrality(G, normalized=True)
b_answer = {"a": 1.0, "b": 0.0, "c": 0.0, "d": 0.0}
for n in sorted(G):
assert b[n] == pytest.approx(b_answer[n], abs=1e-7)
def test_star(self):
"""Betweenness centrality: star """
G=nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b=nx.current_flow_betweenness_centrality(G,normalized=True)
b_answer={'a': 1.0, 'b': 0.0, 'c': 0.0, 'd':0.0}
for n in sorted(G):
assert_almost_equal(b[n],b_answer[n])
def test_star(self):
"Approximate current-flow betweenness centrality: star"
G = nx.Graph()
nx.add_star(G, ["a", "b", "c", "d"])
b = nx.current_flow_betweenness_centrality(G, normalized=True)
epsilon = 0.1
ba = approximate_cfbc(G, normalized=True, epsilon=0.5 * epsilon)
for n in sorted(G):
np.testing.assert_allclose(b[n], ba[n], atol=epsilon)
def mindrevosod(n):
graf = nx.Graph()
graf.add_nodes_from([i for i in range(n)])
graf.add_edge(0, n // 2)
sez1 = [i for i in range(n // 2)]
sez2 = [i for i in range(n // 2, n)]
nx.add_star(graf, sez1)
nx.add_star(graf, sez2)
return graf
def test_star(self):
"Approximate current-flow betweenness centrality: star"
G=nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b=nx.current_flow_betweenness_centrality(G,normalized=True)
epsilon=0.1
ba = approximate_cfbc(G,normalized=True, epsilon=0.5*epsilon)
for n in sorted(G):
assert_allclose(b[n],ba[n],atol=epsilon)
def test_star(self):
"""Betweenness centrality: star """
G = nx.Graph()
nx.add_star(G, ['a', 'b', 'c', 'd'])
b = nx.current_flow_betweenness_centrality_subset(G,
list(G),
list(G),
normalized=True)
b_answer = nx.current_flow_betweenness_centrality(G, normalized=True)
for n in sorted(G):
assert almost_equal(b[n], b_answer[n])
def test_add_star(self):
G = self.G.copy()
nlist = [12, 13, 14, 15]
nx.add_star(G, nlist)
assert_edges_equal(G.edges(nlist), [(12, 13), (12, 14), (12, 15)])
G = self.G.copy()
nx.add_star(G, nlist, weight=2.0)
assert_edges_equal(G.edges(nlist, data=True),
[(12, 13, {'weight': 2.}),
(12, 14, {'weight': 2.}),
(12, 15, {'weight': 2.})])
def test_add_star(self):
G = self.G.copy()
nlist = [12, 13, 14, 15]
nx.add_star(G, nlist)
assert_edges_equal(G.edges(nlist), [(12, 13), (12, 14), (12, 15)])
G = self.G.copy()
nx.add_star(G, nlist, weight=2.0)
assert_edges_equal(G.edges(nlist, data=True),
[(12, 13, {'weight': 2.}),
(12, 14, {'weight': 2.}),
(12, 15, {'weight': 2.})])
def starTopology(G, nodes, number_of_rr, logical_system):
"""
Function used for create Graph of type a star.
"""
nx.add_star(G, nodes)
loopbacks = net.genNet()
#get loopback address for nodes
loaddress = loopbacks.pop(0)
# set loopback addres for RR
nx.set_node_attributes(G, {'R0': {'type':'Route-Reflector', 'loopback': loaddress}})
# set loopback addres for other nodes
[*map(lambda node: nx.set_node_attributes(G, {node: {'loopback':loaddress}}),nodes[1:])]
genEdge(G, nodes, logical_system)
return G
def test_add_star(self):
G = self.G.copy()
nlist = [12, 13, 14, 15]
nx.add_star(G, nlist)
assert edges_equal(G.edges(nlist), [(12, 13), (12, 14), (12, 15)])
G = self.G.copy()
nx.add_star(G, nlist, weight=2.0)
assert edges_equal(
G.edges(nlist, data=True),
[
(12, 13, {
"weight": 2.0
}),
(12, 14, {
"weight": 2.0
}),
(12, 15, {
"weight": 2.0
}),
],
)
G = self.G.copy()
nlist = [12]
nx.add_star(G, nlist)
assert nodes_equal(G, list(self.G) + nlist)
G = self.G.copy()
nlist = []
nx.add_star(G, nlist)
assert nodes_equal(G.nodes, self.Gnodes)
assert edges_equal(G.edges, self.G.edges)
def setUp(self):
self.partitions = []
graph = nx.Graph()
nx.add_star(graph, range(20))
nx.add_path(graph, range(10))
self.partitions.append(NxPartitionGraphBased(graph, representation={node: node % 6 for node in graph}))
graph = nx.DiGraph()
nx.add_star(graph, range(20))
nx.add_path(graph, range(10))
self.partitions.append(NxPartitionGraphBased(graph, representation={node: node % 6 for node in graph}))
self.non_degree_corrected = NewmanReinertNonDegreeCorrected(is_directed=False)
self.degree_corrected = NewmanReinertDegreeCorrected(is_directed=False)
def setUp(self):
self.graphs = []
graph = nx.Graph()
nx.add_path(graph, range(4))
self.graphs.append(graph)
graph = nx.Graph()
nx.add_path(graph, range(10))
self.graphs.append(graph)
graph = nx.DiGraph()
nx.add_star(graph, range(4))
self.graphs.append(graph)
graph = nx.DiGraph()
nx.add_star(graph, range(10))
self.graphs.append(graph)
graph = nx.karate_club_graph()
self.graphs.append(graph)
self.block_sizes = [2, 4, 2, 4, 3]
self.partitions = [
NxPartition(graph,
representation={
node: node % self.block_sizes[i]
for node in graph
}) for i, graph in enumerate(self.graphs)
]
self.representations = [{
node: node % self.block_sizes[i]
for node in graph
} for i, graph in enumerate(self.graphs)]
self.is_directed = [graph.is_directed() for graph in self.graphs]
self.nodes = [len(graph.nodes) for graph in self.graphs]
self.edges = [len(graph.edges) for graph in self.graphs]
def example_networkx():
G = networkx.Graph()
networkx.add_star(G, [0, 1, 2, 3, 4])
networkx.add_path(G, [4, 5, 6, 7, 8])
networkx.add_star(G, [8, 9, 10, 11, 12])
networkx.add_path(G, [6, 13, 14, 15])
networkx.add_star(G, [15, 16, 17, 18])
return G
def build_logo():
shutil.rmtree("docs/_static/logo", ignore_errors=True)
os.makedirs("docs/_static/logo", exist_ok=True)
graph = nx.Graph()
nx.add_path(graph, [0, 1, 2, 3, 4, 5, 0])
nx.add_star(graph, [6, 0, 1, 2, 3, 4, 5])
style = nxv.Style(
graph={
"pad": 1 / 8,
"bgcolor": "#00000000",
"size": "1,1",
"ratio": 1,
},
node=lambda u, d: {
"shape": "circle",
"label": None,
"width": 3 / 4,
"style": "filled",
"fillcolor": "#009AA6" if u % 2 else "#E37222",
"penwidth": 5,
},
edge={"penwidth": 5},
)
path = "docs/_static/logo/logo.svg"
LOGGER.info(f"Rendering {path}")
with open(path, "wb") as f:
f.write(nxv.render(graph, style, algorithm="neato", format="svg"))
for size in [16, 32, 40, 48, 64, 128, 256]:
style_with_dpi = nxv.compose([style, nxv.Style(graph={"dpi": size})])
path = f"docs/_static/logo/logo-{size}.png"
LOGGER.info(f"Rendering {path}")
with open(path, "wb") as f:
f.write(
nxv.render(graph,
style_with_dpi,
algorithm="neato",
format="png"))
def setUp(self):
if self.is_directed:
self.weighted_graph = nx.DiGraph()
else:
self.weighted_graph = nx.Graph()
nx.add_star(self.weighted_graph, range(5))
nx.add_path(self.weighted_graph, range(4, 10))
# add a selfloop
self.weighted_graph.add_edge(4, 4)
self.covariate_graph = self.weighted_graph.copy()
for from_node in self.weighted_graph:
for to_node in self.weighted_graph:
if self.weighted_graph.has_edge(from_node, to_node):
self.weighted_graph[from_node][to_node][
"weight"] = from_node + to_node
self.covariate_graph[from_node][to_node][
"covariate"] = from_node + to_node
representation = {node: node % 3 for node in self.weighted_graph}
self.weighted_partition = self.comparision_class(
self.weighted_graph,
representation=representation,
weighted_graph=True)
self.covariate_partition = self.test_class(
self.covariate_graph,
weighted_graph=False,
with_covariate=True,
representation=representation)
self.unweighted_partition = self.comparision_class(
self.weighted_graph,
representation=representation,
weighted_graph=False)
def remove_edges_from(self, ebunch):
for e in ebunch:
u, v = e[0:2]
self.remove_edge(u, v)
def clear(self):
super().clear()
self.fh.write("Clear graph\n")
G = PrintGraph()
G.add_node("foo")
G.add_nodes_from("bar", weight=8)
G.remove_node("b")
G.remove_nodes_from("ar")
print("Nodes in G: ", G.nodes(data=True))
G.add_edge(0, 1, weight=10)
print("Edges in G: ", G.edges(data=True))
G.remove_edge(0, 1)
G.add_edges_from(zip(range(0, 3), range(1, 4)), weight=10)
print("Edges in G: ", G.edges(data=True))
G.remove_edges_from(zip(range(0, 3), range(1, 4)))
print("Edges in G: ", G.edges(data=True))
G = PrintGraph()
nx.add_path(G, range(10))
nx.add_star(G, range(9, 13))
pos = nx.spring_layout(G, seed=225) # Seed for reproducible layout
nx.draw(G, pos)
plt.show()
def __init__(self):
self.pandemic = nx.Graph()
self.blueCities = ["San Francisco", "Chicago", "Montreal", "New York", "Atlanta", "Washington", "London", "Madrid", "Essen", "Paris", "Milan", "St. Petersburg"]
self.yellowCities = ["Los Angeles", "Mexico City", "Miami", "Bogota", "Lima", "Santiago", "Buenos Aires", "Sao Paulo", "Lagos", "Kinshasa", "Johannesburg", "Khartoum"]
self.blackCities = ["Algiers", "Istanbul", "Cairo", "Moscow", "Baghdad", "Riyadh", "Tehran", "Karachi", "Mumbai", "Delhi", "Chennai", "Kolkata"]
self.redCities = ["Bangkok", "Jakarta", "Beijing", "Shanghai", "Hong Kong", "Ho Chi Minh City", "Seoul", "Taipei", "Manila", "Sydney", "Tokyo", "Osaka"]
self.pandemic.add_nodes_from(self.blueCities, color="blue", cubes=0)
self.pandemic.add_nodes_from(self.yellowCities, color="yellow", cubes=0)
self.pandemic.add_nodes_from(self.blackCities, color="black", cubes=0)
self.pandemic.add_nodes_from(self.redCities, color="red", cubes=0)
nx.add_star(self.pandemic, ["San Francisco", "Tokyo", "Manila", "Los Angeles", "Chicago"])
nx.add_star(self.pandemic, ["Chicago", "San Francisco", "Los Angeles", "Mexico City", "Atlanta", "Montreal"])
nx.add_star(self.pandemic, ["Montreal", "Chicago", "Washington", "New York"])
nx.add_star(self.pandemic, ["New York", "Montreal", "Washington", "London", "Madrid"])
nx.add_star(self.pandemic, ["Atlanta", "Chicago", "Washington", "Miami"])
nx.add_star(self.pandemic, ["Washington", "Miami", "Atlanta", "Montreal", "New York"])
nx.add_star(self.pandemic, ["London", "New York", "Madrid", "Paris", "Essen"])
nx.add_star(self.pandemic, ["Madrid", "Washington", "London", "Paris", "Algiers", "Sao Paulo"])
nx.add_star(self.pandemic, ["Essen", "London", "Paris", "Milan", "St. Petersburg"])
nx.add_star(self.pandemic, ["Paris", "London", "Madrid", "Algiers", "Milan", "Essen"])
nx.add_star(self.pandemic, ["Milan", "Essen", "Paris", "Istanbul"])
nx.add_star(self.pandemic, ["St. Petersburg", "Essen", "Istanbul", "Moscow"])
nx.add_star(self.pandemic, ["Los Angeles", "Sydney", "San Francisco", "Chicago", "Mexico City"])
nx.add_star(self.pandemic, ["Mexico City", "Los Angeles", "Chicago", "Miami", "Bogota", "Lima"])
nx.add_star(self.pandemic, ["Miami", "Atlanta", "Washington", "Bogota", "Mexico City"])
nx.add_star(self.pandemic, ["Bogota", "Mexico City", "Miami", "Sao Paulo", "Buenos Aires", "Lima"])
nx.add_star(self.pandemic, ["Lima", "Mexico City", "Bogota", "Santiago"])
nx.add_star(self.pandemic, ["Santiago", "Lima", "Buenos Aires"])
nx.add_star(self.pandemic, ["Buenos Aires", "Santiago", "Bogota", "Sao Paulo"])
nx.add_star(self.pandemic, ["Sao Paulo", "Bogota", "Buenos Aires", "Madrid", "Lagos"])
nx.add_star(self.pandemic, ["Lagos", "Sao Paulo", "Kinshasa", "Khartoum"])
nx.add_star(self.pandemic, ["Kinshasa", "Lagos", "Johannesburg", "Khartoum"])
nx.add_star(self.pandemic, ["Johannesburg", "Kinshasa", "Khartoum"])
nx.add_star(self.pandemic, ["Khartoum", "Lagos", "Kinshasa", "Johannesburg", "Cairo"])
nx.add_star(self.pandemic, ["Algiers", "Madrid", "Paris", "Istanbul", "Cairo"])
nx.add_star(self.pandemic, ["Istanbul", "Milan", "St. Petersburg", "Moscow", "Baghdad", "Cairo", "Algiers"])
nx.add_star(self.pandemic, ["Cairo", "Algiers", "Istanbul", "Baghdad", "Riyadh"])
nx.add_star(self.pandemic, ["Moscow", "St. Petersburg", "Tehran", "Istanbul"])
nx.add_star(self.pandemic, ["Baghdad", "Istanbul", "Cairo", "Riyadh", "Karachi", "Tehran"])
nx.add_star(self.pandemic, ["Algiers", "Cairo", "Baghdad", "Karachi"])
nx.add_star(self.pandemic, ["Tehran", "Moscow", "Delhi", "Karachi", "Baghdad"])
nx.add_star(self.pandemic, ["Karachi", "Tehran", "Delhi", "Mumbai", "Riyadh", "Baghdad"])
nx.add_star(self.pandemic, ["Mumbai", "Karachi", "Delhi", "Chennai"])
nx.add_star(self.pandemic, ["Delhi", "Tehran", "Karachi", "Mumbai", "Chennai", "Kolkata"])
nx.add_star(self.pandemic, ["Chennai", "Mumbai", "Delhi", "Kolkata", "Bangkok", "Jakarta"])
nx.add_star(self.pandemic, ["Kolkata", "Delhi", "Chennai", "Bangkok", "Hong Kong"])
nx.add_star(self.pandemic, ["Bangkok", "Chennai", "Kolkata", "Hong Kong", "Ho Chi Minh City", "Jakarta"])
nx.add_star(self.pandemic, ["Jakarta", "Chennai", "Bangkok", "Ho Chi Minh City", "Sydney"])
nx.add_star(self.pandemic, ["Beijing", "Seoul", "Shanghai"])
nx.add_star(self.pandemic, ["Shanghai", "Beijing", "Seoul", "Tokyo", "Taipei", "Hong Kong"])
nx.add_star(self.pandemic, ["Hong Kong", "Shanghai", "Taipei", "Manila", "Ho Chi Minh City", "Bangkok", "Kolkata"])
nx.add_star(self.pandemic, ["Ho Chi Minh City", "Bangkok", "Hong Kong", "Manila", "Jakarta"])
nx.add_star(self.pandemic, ["Seoul", "Beijing", "Shanghai", "Tokyo"])
nx.add_star(self.pandemic, ["Taipei", "Bangkok", "Hong Kong", "Manila", "Osaka"])
nx.add_star(self.pandemic, ["Manila", "Taipei", "Hong Kong", "Ho Chi Minh City", "Sydney", "San Francisco"])
nx.add_star(self.pandemic, ["Sydney", "Jakarta", "Manila", "Los Angeles"])
nx.add_star(self.pandemic, ["Tokyo", "Seoul", "Shanghai", "Osaka", "San Francisco"])
nx.add_star(self.pandemic, ["Osaka", "Taipei", "Tokyo"])
#Freezes the board, so no more nodes or edges can be added to the graph
nx.freeze(self.pandemic)
self.outbreakCounter = 0
def remove_edges_from(self, ebunch):
for e in ebunch:
u, v = e[0:2]
self.remove_edge(u, v)
def clear(self):
Graph.clear(self)
self.fh.write("Clear graph\n")
if __name__ == '__main__':
G = PrintGraph()
G.add_node('foo')
G.add_nodes_from('bar', weight=8)
G.remove_node('b')
G.remove_nodes_from('ar')
print("Nodes in G: ", G.nodes(data=True))
G.add_edge(0, 1, weight=10)
print("Edges in G: ", G.edges(data=True))
G.remove_edge(0, 1)
G.add_edges_from(zip(range(0, 3), range(1, 4)), weight=10)
print("Edges in G: ", G.edges(data=True))
G.remove_edges_from(zip(range(0, 3), range(1, 4)))
print("Edges in G: ", G.edges(data=True))
G = PrintGraph()
nx.add_path(G, range(10))
nx.add_star(G, range(9, 13))
nx.draw(G)
plt.show()
G.add_node(1) G.add_nodes_from([3, 4, 5]) # 辺の追加 (頂点も必要に応じて追加されます) G.add_edge(1, 2) G.add_edges_from([(1, 3), (2, 5), (3, 4), (4, 5)]) # 辺の削除 G.remove_edge(3, 4) G.remove_edges_from([(1, 3), (2, 5)]) # 頂点の削除 (削除された頂点に接続されている辺も削除されます) G.remove_node(5) G.remove_nodes_from([3, 4]) # 指定したパス上の頂点と辺を追加 nx.add_path(G, [1, 2, 3, 4, 5]) # 1 → 2 → 3 → 4 → 5 # 指定した閉路上の頂点と辺を追加 nx.add_cycle(G, [1, 2, 3, 4, 5]) # 1 → 2 → 3 → 4 → 5 → 1 # 放射状に頂点と辺を追加 nx.add_star(G, [1, 2, 3, 4, 5]) # 1 → 2, 1 → 3, 1 → 4, 1 → 5 import networkx as nx G = nx.DiGraph() nx.add_path(G, [3, 5, 4, 1, 0, 2, 7, 8, 9, 6]) nx.add_path(G, [3, 0, 6, 4, 2, 7, 1, 9, 8, 5]) nx.nx_agraph.view_pygraphviz(G, prog='fdp') # pygraphvizが必要
def plot(matchers,strings,matcher_names=None,ax=None):
"""
Plots strings and their parent groups for multiple matchers as a graph, with
groups represented as nodes that connect strings
Arguments:
matchers -- a matcher or list of matchers to plot
strings -- a string or list of strings to plot (all connected strings will
also be plotted)
matcher_names -- (optional) a list of strings to label matchers in the plot
legend
ax -- (optional) a matplotlib axis object to draw the plot on.
"""
if isinstance(matchers,Matcher):
matchers = [matchers]
if isinstance(strings,str):
strings = [strings]
if not matcher_names:
matcher_names = [f'matcher{i}' for i in range(len(matchers))]
elif not (len(matcher_names) == len(matchers)):
raise ValueError('matcher_names must be the same length as matchers')
varname = lambda x: f'{x=}'.split('=')[0]
# First build graph representation of the parent groups
G = nx.Graph()
for i,matcher in enumerate(matchers):
m_groups = set(matcher[strings])
for g in m_groups:
group_node = f'{matcher_names[i]}: {g}'
string_nodes = matcher.groups[g]
G.add_nodes_from(string_nodes,type='string',color='w')
if len(string_nodes) > 1:
G.add_nodes_from([group_node],type='group',color=f'C{i}',label=group_node)
nx.add_star(G,[group_node] + string_nodes,color=f'C{i}')
# Now plot graph components in a grid
components = sorted(nx.connected_components(G),key=len,reverse=True)
n_grid = int(np.ceil(np.sqrt(len(components))))
grid_xy = [(x,-y) for y in range(n_grid) for x in range(n_grid)]
if ax is None:
fig, ax = plt.subplots()
for i,component in enumerate(components):
G_sub = G.subgraph(component)
x0,y0 = grid_xy[i]
# Position nodes
if len(component) > 1:
pos = nx.random_layout(G_sub)
pos = nx.kamada_kawai_layout(G_sub,pos=pos,scale=0.25)
pos = {n:(x0+x,y0+y) for n,(x,y) in pos.items()}
else:
pos = {list(component)[0]:(x0,y0)}
edges = list(G_sub.edges(data=True))
edge_coord = [[pos[n0],pos[n1]] for n0,n1,d in edges]
edge_colors = [mplt.colors.to_rgba(d['color']) for n0,n1,d in edges]
lc = mplt.collections.LineCollection(edge_coord,color=edge_colors,zorder=0)
ax.add_collection(lc)
for node,d in G_sub.nodes(data=True):
x,y = pos[node]
if d['type'] == 'group':
ax.scatter(x,y,color=mplt.colors.to_rgb(d['color']),label=d['label'],s=50,zorder=2)
else:
ax.scatter(x,y,color='w',s=200,zorder=1)
ax.text(x,y,node,ha='center',va='center',zorder=3)
ax.axis('off')
legend_handles = [mplt.lines.Line2D([],[],color=mplt.colors.to_rgb(f'C{i}'),markersize=15,marker='.',label=f'{name}')
for i,name in enumerate(matcher_names)]
plt.legend(handles=legend_handles,
bbox_to_anchor=(0.5,-0.2),loc='lower center',
ncol=3,borderaxespad=0)
return ax
