def testCohesiveBlocks2(self):
# Taken from the Moody-White paper
g = Graph.Formula("1-2:3:4:5:6, 2-3:4:5:7, 3-4:6:7, 4-5:6:7, "
"5-6:7:21, 6-7, 7-8:11:14:19, 8-9:11:14, 9-10, "
"10-12:13, 11-12:14, 12-16, 13-16, 14-15, 15-16, "
"17-18:19:20, 18-20:21, 19-20:22:23, 20-21, "
"21-22:23, 22-23")
cbs = g.cohesive_blocks()
self.genericTests(cbs)
expected_blocks = [
list(range(7)),
list(range(23)),
list(range(7)) + list(range(16, 23)),
list(range(6, 16)),
[6, 7, 10, 13],
]
observed_blocks = sorted(
sorted(int(x) - 1 for x in g.vs[bl]["name"]) for bl in cbs)
self.assertEqual(expected_blocks, observed_blocks)
self.assertTrue(cbs.cohesions() == [1, 2, 2, 5, 3])
self.assertTrue(cbs.parents() == [None, 0, 0, 1, 2])
self.assertTrue(
sorted(cbs.hierarchy().get_edgelist()) == [(0, 1), (0,
2), (1,
3), (2,
4)])
def characterize_with_patterns(graph, k):
create_list_neighbors(graph)
global PT
global PS
size_pt_per_k = {2:1, 3:3, 4:9, 5:30}
size_ps_per_k = {2:1, 3:4, 4:15, 5:73}
PT = size_pt_per_k[k]*[0]
PS = []
for v in graph.vs:
PS.append(size_ps_per_k[k]*[0])
v['id_sub'] = -1
for v in graph.vs:
graph_sub = Graph.Formula()
v['id_sub'] = 0
if not 'name' in v.attributes():
v['name'] = str(v.index)
graph_sub.add_vertex(name = v['name'], **{'id_principal' : v.index, 'evol_class' : 1, 'pattern_sub' : 0, 'neighbors_evol_classes' : []})
vext = []
for nei in v['list_neighbors']:
if nei > v.index:
vext.append(graph.vs[nei])
if len(vext) > 0:
extend_subgraph(graph, k, graph_sub, v, vext)
v['id_sub'] = -1
return (PT, PS)
def characterize_with_patterns(self):
self.create_list_neighbors()
self.patterns_tab = 30 * [0]
for v in self._graph.vs:
self.positions_tab.append(73 * [0])
v['id_sub'] = -1
for v in self._graph.vs:
graph_sub = Graph.Formula()
v['id_sub'] = 0
if not 'name' in v.attributes():
v['name'] = str(v.index)
graph_sub.add_vertex(name=v['name'],
**{
'id_principal': v.index,
'evol_class': 1,
'pattern_sub': 0
})
vext = []
for nei in v['list_neighbors']:
if nei > v.index:
vext.append(self._graph.vs[nei])
if len(vext) > 0:
self.extend_subgraph(graph_sub, v, vext)
v['id_sub'] = -1
return (self.patterns_tab, self.positions_tab)
def test_example():
# create a graph consisting of three nodes connected in sequence
# every edge will be upregulating
graph = Graph.Formula('A->B->C')
graph.es['type'] = ['up'] * 2
# the most detrimental node should be B
node = most_detrimental(graph, 'A', 'C')
assert node == 'B'
def import_graph(folder, ego):
home = os.path.expanduser('~')
if not os.path.isfile('%s/GALLERY/%s/%s/Graphs/friends.gml' %
(home, folder, ego)):
return Graph.Formula('')
graph = Graph.Read_GML('%s/GALLERY/%s/%s/Graphs/friends.gml' %
(home, folder, ego))
graph['folder'] = folder
graph['ego'] = ego
return graph
def test_example_other_types():
for edge_types in (('up', 'down'), (0, 1), ('activates', 'inactivates'),
(100, -200)):
# create a graph consisting of three nodes connected in sequence
# every edge will be upregulating
graph = Graph.Formula('A->B->C')
graph.es['type'] = edge_types
# the most detrimental node should be B
node = most_detrimental(graph, 'A', 'C')
assert node == 'B', 'Failed with edge types {} and {}'.format(
*edge_types)
from igraph import Graph
from signatures.networks import get_st_bdd_from_cutsets, get_component_types_dict
from signatures.sig_funs import compute_signatures, print_survival_sig_table
if __name__ == "__main__":
network = Graph.Formula("s -- A:B, A -- C:D, B -- C:E, C -- D:E, D -- t, E -- t")
for v in network.vs:
if v["name"] in ["A", "E"]:
v["compType"] = 1
elif v["name"] in ["B", "C", "D"]:
v["compType"] = 2
bdd, root = get_st_bdd_from_cutsets(network)
component_types = get_component_types_dict(network)
sig_dim_types, survival_sig, system_sig = compute_signatures(bdd, root, component_types)
print_survival_sig_table(survival_sig, sig_dim_types)
def testFormula(self):
tests = [
(None, [], []),
("", [""], []),
("A", ["A"], []),
("A-B", ["A", "B"], [(0, 1)]),
("A --- B", ["A", "B"], [(0, 1)]),
(
"A--B, C--D, E--F, G--H, I, J, K",
["A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K"],
[(0, 1), (2, 3), (4, 5), (6, 7)],
),
(
"A:B:C:D -- A:B:C:D",
["A", "B", "C", "D"],
[(0, 1), (0, 2), (0, 3), (1, 2), (1, 3), (2, 3)],
),
("A -> B -> C", ["A", "B", "C"], [(0, 1), (1, 2)]),
("A <- B -> C", ["A", "B", "C"], [(1, 0), (1, 2)]),
("A <- B -- C", ["A", "B", "C"], [(1, 0)]),
(
"A <-> B <---> C <> D",
["A", "B", "C", "D"],
[(0, 1), (1, 0), (1, 2), (2, 1), (2, 3), (3, 2)],
),
(
"'this is' <- 'a silly' -> 'graph here'",
["this is", "a silly", "graph here"],
[(1, 0), (1, 2)],
),
(
"Alice-Bob-Cecil-Alice, Daniel-Cecil-Eugene, Cecil-Gordon",
["Alice", "Bob", "Cecil", "Daniel", "Eugene", "Gordon"],
[(0, 1), (1, 2), (0, 2), (2, 3), (2, 4), (2, 5)],
),
(
"Alice-Bob:Cecil:Daniel, Cecil:Daniel-Eugene:Gordon",
["Alice", "Bob", "Cecil", "Daniel", "Eugene", "Gordon"],
[(0, 1), (0, 2), (0, 3), (2, 4), (2, 5), (3, 4), (3, 5)],
),
(
"Alice <-> Bob --> Cecil <-- Daniel, Eugene --> Gordon:Helen",
[
"Alice", "Bob", "Cecil", "Daniel", "Eugene", "Gordon",
"Helen"
],
[(0, 1), (1, 0), (1, 2), (3, 2), (4, 5), (4, 6)],
),
(
"Alice -- Bob -- Daniel, Cecil:Gordon, Helen",
["Alice", "Bob", "Daniel", "Cecil", "Gordon", "Helen"],
[(0, 1), (1, 2)],
),
(
'"+" -- "-", "*" -- "/", "%%" -- "%/%"',
["+", "-", "*", "/", "%%", "%/%"],
[(0, 1), (2, 3), (4, 5)],
),
("A-B-C\nC-D", ["A", "B", "C", "D"], [(0, 1), (1, 2), (2, 3)]),
("A-B-C\n C-D", ["A", "B", "C", "D"], [(0, 1), (1, 2), (2, 3)]),
]
for formula, names, edges in tests:
g = Graph.Formula(formula)
self.assertEqual(g.vs["name"], names)
self.assertEqual(g.get_edgelist(), sorted(edges))
def test_survival_signature_values_correct_for_networks(self):
"""
Tests that survival signature is calculated correctly for various network problems.
"""
# Simple series network with single component type.
network = Graph.Formula("s -- 1 -- 2 -- t")
for v in network.vs:
v["compType"] = 1
component_types = get_component_types_dict(network)
bdd, root = get_st_bdd_from_cutsets(network)
sig_dim_types, survival_sig, system_sig = compute_signatures(
bdd, root, component_types)
expected_survival_sig = np.array((0, 0, 1))
expected_system_sig = np.array((1, 0))
self.assertTrue(np.array_equal(survival_sig, expected_survival_sig))
self.assertTrue(np.array_equal(system_sig, expected_system_sig))
'''Network from Figure 1 of "Generalizing the signature to systems with multiple types of components"
by F. Coolen and Coolen-Maturi, 2012'''
network = Graph.Formula(
"s -- 1, 1 -- 2:3, 2 -- 4:5, 3 -- 4:6, 4 -- 5:6, 5 -- t, 6 -- t")
for v in (v for v in network.vs if v["name"] in ["1", "2", "5"]):
v["compType"] = 1
for v in (v for v in network.vs if v["name"] in ["3", "4", "6"]):
v["compType"] = 2
bdd, root = get_st_bdd_from_cutsets(network)
component_types = get_component_types_dict(network)
sig_dim_types, survival_sig, system_sig = compute_signatures(
bdd, root, component_types)
expected_sig = np.array([[0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.1111, 0.3333],
[0.0, 0.0, 0.4444, 0.6667],
[1.0, 1.0, 1.0, 1.0]])
self.assertTrue(
np.array_equal(np.around(survival_sig, decimals=4), expected_sig))
'''Networks from Figures 2 and Figure 4 of the paper
"Bayesian inference for reliability of systems and networks using the survival signature" by L. Aslett et
al, 2015'''
# Reliability network from Fig. 2 and survival signature result from Table 1.
network = Graph.Formula(
"s -- 5:6:7, 5 -- 1, 6 -- 2, 7 -- 3, 1 -- 8, 2 --9, 3 -- 10, 8 -- 4, 9 -- 4,"
"10 -- 4, 4 -- 11, 11 -- t")
for v in (v for v in network.vs if v["name"] in ["1", "2", "3", "4"]):
v["compType"] = 1
for v in (v for v in network.vs
if v["name"] in ["5", "6", "7", "8", "9", "10", "11"]):
v["compType"] = 2
bdd, root = get_st_bdd_from_cutsets(network)
component_types = get_component_types_dict(network)
sig_dim_types, survival_sig, system_sig = compute_signatures(
bdd, root, component_types)
expected_sig = np.zeros((5, 8))
expected_sig[2, 3] = 0.014
expected_sig[2, 4] = 0.057
expected_sig[2, 5] = 0.143
expected_sig[2, 6] = 0.286
expected_sig[2, 7] = 0.500
expected_sig[3, 3] = 0.043
expected_sig[3, 4] = 0.171
expected_sig[3, 5] = 0.393
expected_sig[3, 6] = 0.643
expected_sig[3, 7] = 0.750
expected_sig[4, 3] = 0.086
expected_sig[4, 4] = 0.343
expected_sig[4, 5] = 0.714
expected_sig[4, 6] = 0.857
expected_sig[4, 7] = 1.000
self.assertTrue(
np.array_equal(np.around(survival_sig, decimals=3), expected_sig))
# Reliability network from Fig. 4 and result from Appendix A.
network = Graph.Formula(
"s -- 1:7:9, 1 -- 2:4:5, 9 -- 10, 7 -- 8:10, 2 -- 3, 4 -- 6, 5 -- 6,"
"10 -- 8:11, 8 -- t, 3 -- t, 6 -- t, 11 -- t")
for v in (v for v in network.vs if v["name"] in ["1", "6", "11"]):
v["compType"] = 1
for v in (v for v in network.vs if v["name"] in ["2", "3", "9"]):
v["compType"] = 2
for v in (v for v in network.vs if v["name"] in ["4", "5", "10"]):
v["compType"] = 3
for v in (v for v in network.vs if v["name"] in ["7", "8"]):
v["compType"] = 4
bdd, root = get_st_bdd_from_cutsets(network)
component_types = get_component_types_dict(network)
sig_dim_types, survival_sig, system_sig = compute_signatures(
bdd, root, component_types)
# Test a set of selected values from the signature.
self.assertAlmostEqual(survival_sig[0, 0, 0, 0], 0)
self.assertAlmostEqual(survival_sig[0, 1, 3, 1], 1.0 / 6)
self.assertAlmostEqual(survival_sig[1, 1, 2, 0], 2.0 / 27)
self.assertAlmostEqual(survival_sig[2, 1, 1, 1], 7.0 / 18)
self.assertAlmostEqual(survival_sig[2, 3, 1, 1], 7.0 / 9)
self.assertAlmostEqual(survival_sig[3, 2, 0, 1], 1.0 / 3)
self.assertAlmostEqual(survival_sig[3, 2, 1, 0], 1.0)
