def calculate(self):
new_g1 = DfgMetricUtil.remove_frequencies_from_labels(self.model1)
new_g2 = DfgMetricUtil.remove_frequencies_from_labels(self.model2)
# calulate the nodes similarity first
nodes_metric = DfgNodesSimilarityMetric(self.window,
self.initial_trace,
self.metric_name,
self.model1,
self.model2,
convert=False)
nodes_metric.calculate()
# if the nodes similarity is different than 1
# IPDD removes the different nodes
# then it calculated the edges similarity metric
if nodes_metric.value < 1:
new_g1, new_g2 = DfgMetricUtil.remove_different_nodes(
new_g1, new_g2,
set.union(nodes_metric.diff_added, nodes_metric.diff_removed))
# get the different edges
self.diff_removed = set()
diff_removed = nx.difference(new_g1, new_g2)
for e in diff_removed.edges:
self.diff_removed.add(e)
self.diff_added = set()
diff_added = nx.difference(new_g2, new_g1)
for e in diff_added.edges:
self.diff_added.add(e)
# calculate the edges similarity metric
inter = set(new_g1.edges).intersection(set(new_g2.edges))
self.value = 2 * len(inter) / (len(new_g1.edges) + len(new_g2.edges))
return self.value, self.diff_added, self.diff_removed
def updateCurrentDag(self, dst_prefix, new_activeDag):
"""
:param dst_prefix:
:param new_activeDag: DAG directing any possible router towards dst_prefix
"""
# Get current complete DAG
c_completeDag = self.getCurrentDag(dst_prefix)
# Get current active DAG
c_activeDag = self.getActiveDag(dst_prefix)
# Get edges to set 'active' = False in c_completeDag
edges_to_inactive = nx.difference(c_activeDag, new_activeDag)
# Set them
for (x, y) in edges_to_inactive.edges_iter():
if not c_completeDag.has_edge(x, y):
c_completeDag.add_edge(x, y)
c_completeDag[x][y]['default'] = False
c_completeDag[x][y]['active'] = False
c_completeDag[x][y]['ongoing_flows'] = False
# Get edges to set 'active' = True in c_completeDag
edges_to_active = nx.difference(new_activeDag, c_activeDag)
# Set them
for (x, y) in edges_to_active.edges_iter():
if not c_completeDag.has_edge(x, y):
c_completeDag.add_edge(x, y)
c_completeDag[x][y]['default'] = False
c_completeDag[x][y]['active'] = True
c_completeDag[x][y]['ongoing_flows'] = True
# Set new computed curren complete DAG to dict attribute
self.setCurrentDag(dst_prefix, c_completeDag)
def compare_graphs(nx_graph, cu_graph):
sources, destinations = cu_graph.view_edge_list()
df = pd.DataFrame()
df['source'] = sources.to_pandas()
df['target'] = destinations.to_pandas()
cu_to_nx_graph = nx.from_pandas_edgelist(df, create_using=nx.DiGraph())
# first compare nodes
ds0 = pd.Series(nx_graph.nodes)
ds1 = pd.Series(cu_to_nx_graph.nodes)
if not ds0.equals(ds1):
return False
# second compare edges
diff = nx.difference(nx_graph, cu_to_nx_graph)
if diff.number_of_edges() > 0:
return False
diff = nx.difference(cu_to_nx_graph, nx_graph)
if diff.number_of_edges() > 0:
return False
# need to compare edge weights for weighted graphs as well but currently
# users cannot retrieve edge weights from cugraph Graph objects
# (cugraph issue #319).
return True
def is_same_tree(tree1, tree2):
"""
:param tree1: networkx Graph
:param tree2: networkx Graph
:return: True if the graphs have the same edges, False otherwise
"""
dif1 = nx.difference(tree1, tree2)
dif2 = nx.difference(tree2, tree1)
same_graph = len(dif1.edges) == 0 and len(dif2.edges) == 0
return same_graph
def compare_graphs(nx_graph, cu_graph):
edgelist_df = cu_graph.view_edge_list().reset_index(drop=True)
df = cudf.DataFrame()
df["source"] = edgelist_df["src"]
df["target"] = edgelist_df["dst"]
if len(edgelist_df.columns) > 2:
df["weight"] = edgelist_df["weights"]
cu_to_nx_graph = nx.from_pandas_edgelist(
df.to_pandas(),
source="source",
target="target",
edge_attr=["weight"],
create_using=nx.DiGraph(),
)
else:
cu_to_nx_graph = nx.from_pandas_edgelist(
df.to_pandas(), create_using=nx.DiGraph()
)
# first compare nodes
ds0 = pd.Series(list(nx_graph.nodes)).sort_values(ignore_index=True)
ds1 = pd.Series(list(cu_to_nx_graph.nodes)).sort_values(ignore_index=True)
if not ds0.equals(ds1):
print('ds0 != ds1')
return False
# second compare edges
diff = nx.difference(nx_graph, cu_to_nx_graph)
if diff.number_of_edges() > 0:
print('diff.number_of_edges = ', diff.number_of_edges())
return False
diff = nx.difference(cu_to_nx_graph, nx_graph)
if diff.number_of_edges() > 0:
print('2: diff.number_of_edges = ', diff.number_of_edges())
return False
if len(edgelist_df.columns) > 2:
df0 = cudf.from_pandas(nx.to_pandas_edgelist(nx_graph))
merge = df.merge(df0, on=["source", "target"],
suffixes=("_cugraph", "_nx"))
print("merge = \n", merge)
print(merge[merge.weight_cugraph != merge.weight_nx])
if not merge["weight_cugraph"].equals(merge["weight_nx"]):
print('weights different')
print(merge[merge.weight_cugraph != merge.weight_nx])
return False
return True
def low(G: nx.Graph, root) -> Tuple[Dict[int, int], nx.Graph]:
"""
Calculate low function values for graph nodes.
:param G: input Graph
:param root: root node
:return: tuple (low_map, tree) with dictionary mapping nodes to low function values and tree used in computations
"""
low_map = dict()
T = build_dfs_tree(G, root)
G.remove_nodes_from(set(G.nodes()) - set(T.nodes))
depth = nx.get_node_attributes(T, "depth")
T_nodes = sorted(list(T.nodes()), reverse=True, key=lambda v: depth[v])
T_ud = T.to_undirected()
G_secondary = nx.difference(G, T_ud)
for v in T_nodes:
v_children_low = [low_map[w] for w in T.neighbors(v)]
v_sec_dep = [depth[w] for w in G_secondary.neighbors(v)]
v_dep = depth[v]
low_map[v] = min(v_children_low + v_sec_dep + [v_dep])
return low_map, T
def remove_bcc_edges_from_new_graph(new_graph, bcc_edges_list):
bcc_edges_graph = nx.Graph()
bcc_edges_graph.add_nodes_from(list(new_graph.nodes))
bcc_edges_graph.add_edges_from(bcc_edges_list)
# find the edges that are present in new graph but not in any bcc's.
edge_list = list(nx.difference(new_graph, bcc_edges_graph).edges)
return edge_list
def draw_difference(self, path_old, path_new):
self.draw_path(path_old)
H = self.G.copy()
H.add_edges_from(path_edges(path_new.path))
H_ = nx.difference(self.G, H)
nx.draw(self.G, self.pos)
nx.draw(H_, self.pos, edge_color='blue')
def extract_dodag_before_after(data, list_nodes, neighbors, time, args):
tic = tm.perf_counter()
# Get DODAG some time steps before the anomaly is raised
window_time = args.time_window
data_before = extract_data_up_to(data, time - 1 * window_time, args)
dodag_before = get_dodag(data_before)
# Plot the graph corresponding to the DODAG extracted
# fig, axs = plt.subplots(1, 3, figsize=(30,15))
# axs[0].set_title('DODAG before anomaly')
# nx.draw(dodag_before, with_labels=True, ax=axs[0])
# Get DODAG after the anomaly is raised
data_after = extract_data_up_to(data, time, args)
dodag_after = get_dodag(data_after)
# Plot the graph corresponding to the DODAG extracted
# pos = nx.spring_layout(dodag_after)
# axs[1].set_title('DODAG after anomaly')
# nx.draw(dodag_after, pos=pos, with_labels=True, ax=axs[1])
# nx.draw(dodag_after.subgraph(list_nodes), pos=pos, node_color='orange', with_labels=True, ax=axs[1])
# nx.draw(dodag_after.subgraph(neighbors), pos=pos, node_color='yellow', with_labels=True, ax=axs[1])
# Compute difference between graphs and plot everything
dodag_difference = nx.difference(dodag_after, dodag_before)
dodag_difference.remove_nodes_from(list(nx.isolates(dodag_difference)))
# axs[2].set_title('DODAG difference')
# nx.draw(dodag_difference, with_labels=True, ax=axs[2])
# plt.suptitle(args.scenario)
toc = tm.perf_counter()
# print('Everything DODAG took {:.5f}'.format(toc - tic))
# plt.show()
if (len(dodag_difference) == 0):
return False, []
nodes_changing = dodag_difference.nodes()
return True, nodes_changing
def cal_existed_edges_ratio(g1, g2, type):
g1.remove_node(keyword)
g2.remove_node(keyword)
ngs = set(g1.nodes())
ngs1 = set(g2.nodes())
ngs2 = ngs & ngs1
# 计算每个月的网络
ngs = ngs - ngs2
init_edge_num = len(g1.edges())
for n in ngs:
g1.remove_node(n)
extra_node_edge = init_edge_num - len(g1.edges())
# 计算总网络
ngs1 = ngs1 - ngs2
for n in ngs1:
g2.remove_node(n)
g3 = nx.difference(g1, g2)
# 计算保留率
r = 1 - (len(g3.edges()) + extra_node_edge) / init_edge_num
print(r)
return r
def write_flowtable(self, f):
output = {}
graph = nx.Graph(self.graph)
mst = nx.minimum_spanning_tree(graph)
# Links that need to be disabled
diff = nx.difference(graph, mst)
for switch in self.switches.values():
output[switch.dpid] = {}
output[switch.dpid]["fdb"] = {}
# Write FDB
for src in switch.fdb._fdb.keys():
output[switch.dpid]["fdb"][src.mac] = {}
for dst, nex in switch.fdb._fdb[src].items():
port = self.graph[switch.name][nex.name]["port"]
output[switch.dpid]["fdb"][src.mac][dst.mac] = port
# Write list of blocked ports (to remove loops)
output[switch.dpid]["blocked_ports"] = [
self.graph[switch.name][u]["port"]
for u in diff.neighbors_iter(switch.name)
]
yaml.dump(output, f, explicit_start=True, default_flow_style=False)
def difference(graph2, graph1):
D1 = nx.difference(graph2, graph1)
pos = nx.circular_layout(D1)
D2 = nx.difference(graph1, graph2) # edges in graph1 but not in graph2
pos = nx.circular_layout(D2)
nx.draw_networkx_nodes(D1,pos, node_color="g", node_size=1000)
nx.draw_networkx_edges(D1,pos, edge_color='g',width=10)
nx.draw_networkx_nodes(D2,pos, node_color="r", node_size=1000)
nx.draw_networkx_edges(D2,pos, edge_color='r',width=10)
plt.show()
# plt.savefig("M.PDF",facecolor='pink') #save graph
return difference
def difference(self, other_kg):
""" Returns the difference of the knowledge graph and other_kg
:param other_kg: The other knowledge graph
:type other_kg: :class:`.KnowledgeGraph`
"""
# Returns a KnowledgeGraph containing edges that exist in self but not in other_kg
return KnowledgeGraph(nx.difference(self.net, other_kg.net))
def test_pbc_on_off_difference(universe, lipid_heads):
lfls_pbc_on = LeafletFinder(universe, lipid_heads, cutoff=7, pbc=True)
lfls_pbc_off = LeafletFinder(universe, lipid_heads, cutoff=7, pbc=False)
pbc_on_graph = lfls_pbc_on.graph
pbc_off_graph = lfls_pbc_off.graph
diff_graph = NX.difference(pbc_on_graph, pbc_off_graph)
assert_equal(set(diff_graph.edges), {(69, 153), (73, 79), (206, 317),
(313, 319)})
def test_difference(testgraph):
"""
Test the Difference of the two graphs are same
"""
a = nx.difference(testgraph[0], testgraph[1])
b = sg.digraph_operations.difference(testgraph[2], testgraph[3])
digraph_equals(a, b)
def test_difference(testgraph):
"""
Test the Difference of the two graphs are same
"""
a = nx.difference(testgraph[0], testgraph[1])
b = sg.graph_operations.difference(testgraph[2], testgraph[3])
graph_equals(a, b)
def __eq__(self, other: 'ScoringGraph'):
if other == None:
return False
if (set(self.node.keys()) == set(other.node.keys())):
result = nx.difference(self, other)
return len(result) == 0
else:
return False
def getchangedlinks(G, Gnew):
#get all the changed links
edges_add = []
Gdiff = nx.difference(Gnew, G)
edges_add.append(Gdiff.edges())
edges_rm = []
Gdiff = nx.difference(G, Gnew)
edges_rm.append(Gdiff.edges())
# pdb.set_trace()
# for e in edges:
# nodes.append(e[0])
# nodes.append(e[1])
return edges_add, edges_rm
def main():
print("Processing Aqualung Wiki")
Aq = Aqualung()
print("Processing Ian Anderson Wiki")
IanA = Ian()
print("Generating Similarity rankings for Aqualung")
As = nx.simrank_similarity(Aq)
A = [[As[u][v] for v in sorted(As[u])] for u in sorted(As)]
sim_array = array(A)
print("Generating Similarity rankings for Ian Anderson")
Ia = nx.simrank_similarity(Aq)
IanAr = [[Ia[u][v] for v in sorted(Ia[u])] for u in sorted(Ia)]
Ian_array = array(IanAr)
AqL = list(Aq.nodes)
IanAL = list(IanA.nodes)
print("\nSimilarities for Aqualung")
for x in range(len(sim_array)):
for y in range(len(sim_array[x])):
if x == y:
break
elif sim_array[x][y] >= 0.01:
print(AqL[x], " | ", AqL[y], " | ", sim_array[x][y])
print("\nSimilarities for Ian Anderson")
for x in range(len(Ian_array)):
for y in range(len(Ian_array[x])):
if x == y:
break
elif Ian_array[x][y] >= 0.01:
print(IanAL[x], " | ", IanAL[y], " | ", Ian_array[x][y])
#removing nodes for networkx.difference
for node in AqL:
if (node in IanAL):
continue
else:
Aq.remove_node(node)
for node in IanAL:
if (node in AqL):
continue
else:
IanA.remove_node(node)
print("\nComputing Difference")
D = nx.difference(Aq, IanA)
D.remove_nodes_from(list(nx.isolates(D)))
print(nx.info(D))
#Computing Intersection
print("\nIntersection")
I = nx.intersection(Aq, IanA)
I.remove_nodes_from(list(nx.isolates(I)))
print(nx.info(I))
def compare_graphs(nx_graph, cu_graph):
sources, destinations, values = cu_graph.view_edge_list()
df = cudf.DataFrame()
df['source'] = sources
df['target'] = destinations
if values is not None:
df['weight'] = values
cu_to_nx_graph = nx.from_pandas_edgelist(df.to_pandas(),
source='source',
target='target',
edge_attr=['weight'],
create_using=nx.DiGraph())
else:
cu_to_nx_graph = nx.from_pandas_edgelist(df.to_pandas(),
create_using=nx.DiGraph())
# first compare nodes
ds0 = pd.Series(nx_graph.nodes)
ds1 = pd.Series(cu_to_nx_graph.nodes)
if not ds0.equals(ds1):
return False
# second compare edges
diff = nx.difference(nx_graph, cu_to_nx_graph)
if diff.number_of_edges() > 0:
return False
diff = nx.difference(cu_to_nx_graph, nx_graph)
if diff.number_of_edges() > 0:
return False
if values is not None:
df0 = cudf.from_pandas(nx.to_pandas_edgelist(nx_graph))
df0 = df0.sort_values(by=['source', 'target'])
df1 = df.sort_values(by=['source', 'target'])
if not df0['weight'].equals(df1['weight']):
return False
return True
def test_difference():
G = nx.Graph()
H = nx.Graph()
G.add_nodes_from([1, 2, 3, 4])
G.add_edge(1, 2)
G.add_edge(2, 3)
H.add_nodes_from([1, 2, 3, 4])
H.add_edge(2, 3)
H.add_edge(3, 4)
D = nx.difference(G, H)
assert set(D.nodes()) == set([1, 2, 3, 4])
assert sorted(D.edges()) == [(1, 2)]
D = nx.difference(H, G)
assert set(D.nodes()) == set([1, 2, 3, 4])
assert sorted(D.edges()) == [(3, 4)]
D = nx.symmetric_difference(G, H)
assert set(D.nodes()) == set([1, 2, 3, 4])
assert sorted(D.edges()) == [(1, 2), (3, 4)]
def test_difference2():
G = nx.Graph()
H = nx.Graph()
G.add_nodes_from([1, 2, 3, 4])
H.add_nodes_from([1, 2, 3, 4])
G.add_edge(1, 2)
H.add_edge(1, 2)
G.add_edge(2, 3)
D = nx.difference(G, H)
assert set(D.nodes()) == {1, 2, 3, 4}
assert sorted(D.edges()) == [(2, 3)]
D = nx.difference(H, G)
assert set(D.nodes()) == {1, 2, 3, 4}
assert sorted(D.edges()) == []
H.add_edge(3, 4)
D = nx.difference(H, G)
assert set(D.nodes()) == {1, 2, 3, 4}
assert sorted(D.edges()) == [(3, 4)]
def test_difference2():
G = nx.Graph()
H = nx.Graph()
G.add_nodes_from([1, 2, 3, 4])
H.add_nodes_from([1, 2, 3, 4])
G.add_edge(1, 2)
H.add_edge(1, 2)
G.add_edge(2, 3)
D = nx.difference(G, H)
assert_equal(set(D.nodes()), set([1, 2, 3, 4]))
assert_equal(sorted(D.edges()), [(2, 3)])
D = nx.difference(H, G)
assert_equal(set(D.nodes()), set([1, 2, 3, 4]))
assert_equal(sorted(D.edges()), [])
H.add_edge(3, 4)
D = nx.difference(H, G)
assert_equal(set(D.nodes()), set([1, 2, 3, 4]))
assert_equal(sorted(D.edges()), [(3, 4)])
def test_difference():
G=nx.Graph()
H=nx.Graph()
G.add_nodes_from([1,2,3,4])
G.add_edge(1,2)
G.add_edge(2,3)
H.add_nodes_from([1,2,3,4])
H.add_edge(2,3)
H.add_edge(3,4)
D=nx.difference(G,H)
assert_equal( set(D.nodes()) , set([1,2,3,4]) )
assert_equal( sorted(D.edges()) , [(1,2)] )
D=nx.difference(H,G)
assert_equal( set(D.nodes()) , set([1,2,3,4]) )
assert_equal( sorted(D.edges()) , [(3,4)] )
D=nx.symmetric_difference(G,H)
assert_equal( set(D.nodes()) , set([1,2,3,4]) )
assert_equal( sorted(D.edges()) , [(1,2),(3,4)] )
def compare_graphs(nx_graph, cu_graph):
edgelist_df = cu_graph.view_edge_list()
df = cudf.DataFrame()
df['source'] = edgelist_df['src']
df['target'] = edgelist_df['dst']
if len(edgelist_df.columns) > 2:
df['weight'] = edgelist_df['weights']
cu_to_nx_graph = nx.from_pandas_edgelist(df.to_pandas(),
source='source',
target='target',
edge_attr=['weight'],
create_using=nx.DiGraph())
else:
cu_to_nx_graph = nx.from_pandas_edgelist(df.to_pandas(),
create_using=nx.DiGraph())
# first compare nodes
ds0 = pd.Series(nx_graph.nodes)
ds1 = pd.Series(cu_to_nx_graph.nodes)
if not ds0.equals(ds1):
return False
# second compare edges
diff = nx.difference(nx_graph, cu_to_nx_graph)
if diff.number_of_edges() > 0:
return False
diff = nx.difference(cu_to_nx_graph, nx_graph)
if diff.number_of_edges() > 0:
return False
if len(edgelist_df.columns) > 2:
df0 = cudf.from_pandas(nx.to_pandas_edgelist(nx_graph))
df0 = df0.sort_values(by=['source', 'target'])
df1 = df.sort_values(by=['source', 'target'])
if not df0['weight'].equals(df1['weight']):
return False
return True
def diff_graphs0(self, g1, g2):
""" Diff nx graphs. """
g1_copy = g1.copy()
g1_copy.remove_nodes_from(n for n in g1 if not n in g2)
g2_copy = g2.copy()
g2_copy.remove_nodes_from(n for n in g2 if not n in g1)
g1_minus_g2 = nx.difference(g1_copy, g2_copy)
#self.print_graph (g1_minus_g2)
return g1_minus_g2
def is_different(dag1, dag2):
diff_dag = None
is_diff = False
try:
diff_dag = nx.difference(dag1, dag2)
is_diff = len(diff_dag.edges()) != 0
except nx.exception.NetworkXError:
is_diff = True
return is_diff
def main():
print("Generating Aq")
Aq = Aqualung()
print("Generating IanA")
IanA = Ian()
print("Computing similarity")
nx.write_edgelist(Aq, "aq.edgelist")
#As = nx.simrank_similarity(Aq,max_iterations=5)
#print("Done with simrank")
G1 = nx.DiGraph()
G1.add_nodes_from([1, 2, 3, 4, 5])
G1.add_edges_from([(1, 2), (1, 3), (1, 4), (4, 5)])
#s = nx.simrank_similarity(G1)
#A = [[s[u][v] for v in sorted(s[u])] for u in sorted(s)]
#sim_array = array(A)
#print(sim_array)
#AqL = list(Aq.nodes)
#IanAL = list(IanA.nodes)
#for node in AqL:
# for node2 in AqL:
# print("Checking",node," ",node2)
# sim_val = nx.simrank_similarity(Aq,source=node,target=node2)
# if sim_val >= 0.01:
# print(node, node2, sim_val)
#removing nodes for networkx.difference
AqL = list(Aq.nodes)
IanAL = list(IanA.nodes)
for node in AqL:
if (node in IanAL):
continue
else:
Aq.remove_node(node)
for node in IanAL:
if (node in AqL):
continue
else:
IanA.remove_node(node)
print("\nComputing Difference")
D = nx.difference(Aq, IanA)
D.remove_nodes_from(list(nx.isolates(D)))
print(nx.info(D))
#Computing Intersection
print("\nIntersection")
I = nx.intersection(Aq, IanA)
I.remove_nodes_from(list(nx.isolates(I)))
print(nx.info(I))
def calcConvergeTime(self, t1, t2, converge_time = -1):
# calculate the exact time that the network becomes stable:
# switch states and controller states match with the physical network
converged = True
# controllers should be consistent with each other and the underlying graph
for c in self.controller_names:
ctrl = self.objs[c]
if not nx.is_isomorphic(self.graph, ctrl.graph):
converged = False
cgraph = nx.Graph(ctrl.graph)
node_diff = set(self.graph.nodes()) - set(cgraph.nodes())
cgraph.add_nodes_from(list(node_diff))
node_diff = set(cgraph.nodes()) - set(self.graph.nodes())
self.graph.add_nodes_from(list(node_diff))
diff = nx.difference(self.graph, cgraph)
print "%f %s not converged %s"%(self.ctx.now, ctrl.name, diff.edges())
#break
else:
print "%f %s converged"%(self.ctx.now, ctrl.name)
# switches should also be consistent with each other and the controllers
if converged:
c = self.objs[self.controller_names[0]]
rules = {}
for s in self.switch_names:
rules[s] = {}
sp = nx.shortest_paths.all_pairs_shortest_path(c.graph)
for host in c.hosts:
for h2 in c.hosts:
if h2 == host:
continue
if h2.name in sp[host.name]:
p = SourceDestinationPacket(host.address, h2.address)
path = zip(sp[host.name][h2.name], \
sp[host.name][h2.name][1:])
for (a, b) in path[1:]:
link = c.graph[a][b]['link']
rules[a][p.pack()] = link
for s in self.switch_names:
sw = self.objs[s]
if rules[s] != sw.rules:
print self.ctx.now, s, len(rules), len(sw.rules)
converged = False
break
if converged:
if converge_time == -1:
print "CONVERGE_TIME: ", t1, t2, self.ctx.now
self.ctx.schedule_task(10, lambda: self.calcConvergeTime(t1, t2, self.ctx.now))
else:
self.ctx.schedule_task(10, lambda: self.calcConvergeTime(t1, t2, converge_time))
else:
self.ctx.schedule_task(10, lambda: self.calcConvergeTime(t1, t2, -1))
def test_save_results(self):
se1 = StructureConstraintBasedEstimator(self.s1, 0.1, 0.1)
se1.ctpc_algorithm()
file_name = './PyCTBN/tests/estimators/test_save.json'
se1.save_results(file_name)
with open(file_name) as f:
js_graph = json.load(f)
result_graph = nx.json_graph.node_link_graph(js_graph)
self.assertFalse(
nx.difference(se1._complete_graph, result_graph).edges)
os.remove(file_name)
def main(argv):
args = get_args(argv)
inf_gene_pairs = np.genfromtxt(args['reg_net_file'],
dtype=str,
delimiter='\t')
large_known_net = np.genfromtxt(args['known_network_file'],
dtype=str,
delimiter='\t')
# get set of gene names that are in both inferred and known networks
inf_gene_set = get_inf_gene_set(inf_gene_pairs)
kno_gene_set = set(
large_known_net[:, (0, 2)].flat) # 0th an 2nd col are gene names
gene_set = inf_gene_set.intersection(kno_gene_set)
inf_graph = get_inf_graph(inf_gene_pairs, gene_set, 0.776)
kno_graph = get_kno_graph(large_known_net, gene_set)
# print number of nodes and edges in inferred and known regulatory networks
print str(len(inf_graph.nodes())) + ' nodes in inferred graph'
print str(len(kno_graph.nodes())) + ' nodes in known graph'
print str(len(inf_graph.edges())) + ' edges in inferred graph'
print str(len(kno_graph.edges())) + ' edges in known graph'
# find edges exclusive to each graph and shared between graphs
inf_graph_only = nx.difference(inf_graph, kno_graph)
kno_graph_only = nx.difference(kno_graph, inf_graph)
shared_graph = nx.difference(inf_graph, inf_graph_only)
# print edges shared between graphs
print str(len(inf_graph_only.edges())) + ' of ' + str(
len(inf_graph.edges())) + ' edges unique to inf_graph'
print str(len(kno_graph_only.edges())) + ' of ' + str(
len(kno_graph.edges())) + ' edges unique to kno_graph'
print 'the ' + str(len(inf_graph.edges()) -
len(inf_graph_only.edges())) + ' edge(s) in common are:'
print shared_graph.edges()
# plot ego graph for specific gene
write_ego_graph(shared_graph, EGO_GRAPH_GENE)
def difference(cls, topo1, topo2):
try:
self = cls(nx.difference(topo1, topo2))
except nx.NetworkXError:
return None
if len(self.edges()) == 0:
return None
self.copy_attributes(topo1)
self.reconcile_attributes(topo1)
return self
def difference(cls,topo1,topo2):
try:
self = cls(nx.difference(topo1,topo2))
except nx.NetworkXError:
return None
if len(self.edges()) == 0:
return None
self.copy_attributes(topo1)
self.reconcile_attributes(topo1)
return self
def spurious_edges(self) -> typing.List:
"""Return the spurious edges present in the estimated structure, if a prior net structure is present in
``_sample_path.structure``.
:return: A list containing the spurious edges
:rtype: List
"""
if not self._sample_path.has_prior_net_structure:
return []
real_graph = nx.DiGraph()
real_graph.add_nodes_from(self._sample_path.structure.nodes_labels)
real_graph.add_edges_from(self._sample_path.structure.edges)
return nx.difference(real_graph, self._complete_graph).edges
def test_difference_multigraph_attributes():
g = nx.MultiGraph()
g.add_edge(0, 1, key=0)
g.add_edge(0, 1, key=1)
g.add_edge(0, 1, key=2)
h = nx.MultiGraph()
h.add_edge(0, 1, key=0)
h.add_edge(0, 1, key=3)
gh = nx.difference(g, h)
assert set(gh.nodes()) == set(g.nodes())
assert set(gh.nodes()) == set(h.nodes())
assert sorted(gh.edges()) == [(0, 1), (0, 1)]
assert sorted(gh.edges(keys=True)) == [(0, 1, 1), (0, 1, 2)]
def test_difference_multigraph_attributes():
g = nx.MultiGraph()
g.add_edge(0, 1, key=0)
g.add_edge(0, 1, key=1)
g.add_edge(0, 1, key=2)
h = nx.MultiGraph()
h.add_edge(0, 1, key=0)
h.add_edge(0, 1, key=3)
gh = nx.difference(g, h)
assert_equal(set(gh.nodes()), set(g.nodes()))
assert_equal(set(gh.nodes()), set(h.nodes()))
assert_equal(sorted(gh.edges()), [(0, 1)])
assert_equal(sorted(gh.edges(keys=True)), [(0, 1, 3)])
def bayesnet_compare_plot(bayesmodel_true, bayesmodel_predicted, ax, pos):
nx.draw_networkx_nodes(bayesmodel_true,
ax=ax,
pos=pos,
node_size=800,
node_color='g',
alpha=0.6)
nx.draw_networkx_labels(bayesmodel_true,
ax=ax,
pos=pos,
font_size=16,
font_weight='bold')
nx.draw_networkx_edges(nx.intersection(bayesmodel_true,
bayesmodel_predicted),
ax=ax,
pos=pos,
width=4,
edge_color='k',
style='solid',
alpha=0.6,
arrowsize=25)
nx.draw_networkx_edges(nx.difference(bayesmodel_true,
bayesmodel_predicted),
ax=ax,
pos=pos,
width=4,
edge_color='r',
style='dashed',
alpha=0.6,
arrowsize=25)
nx.draw_networkx_edges(nx.difference(bayesmodel_predicted,
bayesmodel_true),
ax=ax,
pos=pos,
width=4,
edge_color='b',
style='dashed',
alpha=0.6,
arrowsize=25)
def test_difference_multigraph_attributes():
g = nx.MultiGraph()
g.add_edge(0, 1, key=0)
g.add_edge(0, 1, key=1)
g.add_edge(0, 1, key=2)
h = nx.MultiGraph()
h.add_edge(0, 1, key=0)
h.add_edge(0, 1, key=3)
gh = nx.difference(g, h)
assert_equal( set(gh.nodes()) , set(g.nodes()) )
assert_equal( set(gh.nodes()) , set(h.nodes()) )
assert_equal( sorted(gh.edges()) , [(0,1)] )
assert_equal( sorted(gh.edges(keys=True)) , [(0,1,3)] )
def test_difference_attributes():
g = nx.Graph()
g.add_node(0, x=4)
g.add_node(1, x=5)
g.add_edge(0, 1, size=5)
g.graph['name'] = 'g'
h = g.copy()
h.graph['name'] = 'h'
h.graph['attr'] = 'attr'
h.node[0]['x'] = 7
gh = nx.difference(g, h)
assert_equal( set(gh.nodes()) , set(g.nodes()) )
assert_equal( set(gh.nodes()) , set(h.nodes()) )
assert_equal( sorted(gh.edges()) , [])
h.remove_node(0)
assert_raises(nx.NetworkXError, nx.intersection, g, h)
def removeAllocationEntry(self, prefix, flow):
"""
"""
# Wait until flow finishes
time.sleep(flow['duration'])
# Acquire locks for self.flow_allocation and self.dags
# dictionaries
self.flowAllocationLock.acquire()
self.dagsLock.acquire()
log.info(lineend)
if prefix not in self.flow_allocation.keys():
# prefix not in table
raise KeyError("The is no such prefix allocated: %s"%str(prefix))
else:
if flow in self.flow_allocation[prefix].keys():
path_list = self.flow_allocation[prefix].pop(flow, None)
else:
raise KeyError("%s is not alloacated in this prefix %s"%str(repr(flow)))
t = time.strftime("%H:%M:%S", time.gmtime())
log.info("%s - Flow REMOVED from Paths\n"%t)
log.info("\t* Dest_prefix: %s\n"%prefix)
to_log = "\t* Paths (%s): %s\n"
log.info(to_log%(len(path_list), str(self.toLogRouterNames(path_list))))
log.info("\t* Flow: %s\n"%self.toLogFlowNames(flow))
# Current dag for destination
current_dag = self.getCurrentDag(prefix)
# Get the active Dag
activeDag = self.getActiveDag(prefix)
# Log active DAG
to_log = "\t* removeAllocationEntry: current active DAG\n\t %s\n"
log.info(to_log%str(self.toLogDagNames(activeDag).edges()))
# Get current remaining allocated flows for destination
remaining_flows = self.getAllocatedFlows(prefix)
# Create DAG showing remaining flow paths only
edges_with_flows = []
for (f, pl) in remaining_flows:
edges_with_flows += self.getEdgesFromPathList(pl)
edges_with_flows = list(set(edges_with_flows))
remaining_traffic_dag = nx.DiGraph()
remaining_traffic_dag.add_nodes_from(activeDag.nodes())
remaining_traffic_dag.add_edges_from(edges_with_flows)
# Difference with active DAG can be set to 'ongoing_flows' = False
to_set_noflows = nx.difference(activeDag, remaining_traffic_dag)
for (x, y) in to_set_noflows.edges_iter():
current_dag[x][y]['ongoing_flows'] = False
# Set the new calculated dag to its destination prefix dag
self.setCurrentDag(prefix, current_dag)
# Check if we can set destination forwarding to the initial
# default OSPF DAG
if len(remaining_flows) == 0:
# Log a bit
to_log = "\t* No more flows remain to prefix."
to_log += " Re-setting to initial OSPF DAG\n"
log.info(to_log)
# Set forwarding to original
self.setOSPFOriginalDAG(prefix)
else:
# Log it only
log.info("\t* Some flows to prefix still remain.\n")
# Log final DAG that is foced
activeDag = self.getActiveDag(prefix)
to_log = "\t* removePrefixLies: final active DAG\n\t %s\n"
log.info(to_log%str(self.toLogDagNames(activeDag).edges()))
# Force it to fibbing
self.sbmanager.add_dag_requirement(prefix, activeDag.copy())
log.info(lineend)
# Release locks
self.flowAllocationLock.release()
self.dagsLock.release()
def run(self, raw_data, objects):
graphs = []
new_objects = []
for overlay in self.__sub_overlays:
g = overlay.substitutes(raw_data)
graphs.append(g)
graph = graphs.pop(0)
for h in graphs:
graph = nx.compose(graph, h)
components = nx.connected_component_subgraphs(graph)
closed_polygons = []
polygons = []
for component in components:
minimum_cycles = planar_cycles(component)
collected_options = itertools.chain(*nx.get_node_attributes(component,'options').values())
options = list(set(collected_options))
# the polygons are the same as the minimum cycles
closed_polygons += minimum_cycles
path_graph = nx.Graph()
path_graph.add_nodes_from(component.nodes())
for polygon in minimum_cycles:
polygons.append(Polygon(polygon, options))
path_graph.add_cycle(polygon)
remaining_graph = nx.difference(component, path_graph)
for n in remaining_graph.nodes():
if remaining_graph.degree(n) == 0:
remaining_graph.remove_node(n)
if len(remaining_graph.edges()) > 0:
remaining_components = nx.connected_component_subgraphs(remaining_graph)
for c in remaining_components:
new_objects.append(OpenGraph(c, options))
new_objects = new_objects + polygons
z_order_graph = nx.DiGraph()
z_order_graph.add_nodes_from(new_objects)
for i in range(0, len(polygons)):
polygon1 = polygons[i]
for j in range(i + 1, len(polygons)):
polygon2 = polygons[j]
if polygon1 != polygon2:
if polygon1.contains(polygon2):
z_order_graph.add_edge(polygon1, polygon2)
elif polygon2.contains(polygon1):
z_order_graph.add_edge(polygon2, polygon1)
for obj in new_objects:
for edge in obj.edges():
if 'below' in graph[edge[0]][edge[1]]:
below = graph[edge[0]][edge[1]]['below']
if below != None:
for obj_above in new_objects:
if obj != obj_above:
if obj_above.has_edge(below.start(), below.end()):
z_order_graph.add_edge(obj, obj_above)
if 'above' in graph[edge[0]][edge[1]]:
above = graph[edge[0]][edge[1]]['above']
if above != None:
for obj_below in new_objects:
if obj != obj_below:
if obj_below.has_edge(above.start(), above.end()):
z_order_graph.add_edge(obj_below, obj)
if 'z_order' in graph[edge[0]][edge[1]]:
z_order = graph[edge[0]][edge[1]]['z_order']
if z_order != None:
for other_obj in new_objects:
if obj != other_obj:
if (isinstance(other_obj, Polygon) and other_obj.frame().intersects(edge)) or (isinstance(other_obj, OpenGraph) and other_obj.intersects(edge)):
if z_order == 'above':
z_order_graph.add_edge(other_obj, obj)
elif z_order == 'below':
z_order_graph.add_edge(obj, other_obj)
else:
raise ValueError, "Wrong value for z_order."
cycle_gen = nx.simple_cycles(z_order_graph)
try:
cycles = list(cycle_gen)
for cycle in cycles:
cycle_edges = [(cycle[i], cycle[i + 1]) for i in range(0, len(cycle) - 1)]
for edge in cycle_edges:
z_order_graph.remove_edge(edge[0], edge[1])
if cycles:
warnings.warn("The diagram contains objects. that have an ambiguous z-order. Shaape estimates their z-order.", RuntimeWarning)
except:
pass
current_z_order = 0
while z_order_graph.nodes():
nodes_without_predecessors = [node for node in z_order_graph.nodes() if not z_order_graph.predecessors(node)]
for node in nodes_without_predecessors:
node.set_z_order(current_z_order)
current_z_order = current_z_order + 1
z_order_graph.remove_nodes_from(nodes_without_predecessors)
for o in new_objects:
if type(o) == Polygon or type(o) == OpenGraph:
o.reduce_nodes()
objects = objects + new_objects
objects.append(graph)
self._objects = objects
self._parsed_data = raw_data
return
def test_difference_raise():
G = nx.path_graph(4)
H = nx.path_graph(3)
GH = nx.difference(G, H)
G1.edges() """ Intersection de graphes """ # Les noeuds de H et G doivent etre les memes H.clear() H.add_nodes_from([1,2,3,4]) H.add_edge(1,2) GH4=nx.intersection(G,H) GH4.nodes() #[1,2,3,4] GH4.edges() #[(1,2)] """ Difference de graphes """ # Les noeuds de H et G doivent etre les memes GH5=nx.difference(G,H) GH5.nodes() # [1,2,3,4] GH5.edges() # [((2,3),(3,4)] # Retourne un graphe avec des aretes qui existent dans G mais pas dans H """ Difference symetrique de graphes """ # Les noeuds de H et G doivent etre les memes GH6=nx.symmetric_difference(G,H) GH6.nodes() # [1,2,3,4] GH6.edges() # [((2,3),(3,4)] # Retourne un graphe avec des aretes qui sont soit dans G soit dans H mais pas les deux
G_list['firstth'] = G6
G7 = nx.read_gml("hdb_collab_cluster_sample_firstth20.gml")
G_list['firstth20'] = G7
print 'ground truth sample cluster:'
print 'nodes:', len(G0.nodes())
print 'edges:', len(G0.edges())
print '\n\n'
for node in G1.nodes():
for key, value in G_list.iteritems():
if not value.has_node(node):
value.add_node(node)
for key, value in G_list.iteritems():
print key, 'sample cluster'
print 'nodes:', len(value.nodes())
print 'edges:', len(value.edges())
print '\nDIFF normal test:'
print 'missing edges:', len(nx.difference(G1, value).edges())
print 'added edges:', len(nx.difference(value, G1).edges())
print '\n\n'
def test_mixed_type_difference():
G = nx.Graph()
H = nx.MultiGraph()
U = nx.difference(G,H)