def test_randomize_edges_by_switching(self):
for i in range(10):
n = random.randint(1, 10)
G = Network.generate_random(n_vertices=n, indegree_bounds=[1, 3])
G_tag = G.copy()
G_tag.randomize_edges_by_switching(include_self_loops=False)
for v, v_tag in zip(G.vertices, G_tag.vertices):
self.assertTrue(v.function == v_tag.function)
self.assertTrue(
len(v.predecessors()) == len(v_tag.predecessors()))
G = Network(vertex_names=["A", "B", "C"],
edges=[("A", "B"), ("A", "A"), ("B", "A"), ("B", "C")],
vertex_functions=[None] * 3)
for i in range(10):
G_tag = G.copy()
G_tag.randomize_edges_by_switching(include_self_loops=False)
self.assertTrue(
G_tag.get_vertex("A").predecessors() ==
[G_tag.get_vertex("A"),
G_tag.get_vertex("B")])
self.assertTrue(
G_tag.get_vertex("B").predecessors() ==
[G_tag.get_vertex("A")])
self.assertTrue(
G_tag.get_vertex("C").predecessors() ==
[G_tag.get_vertex("B")])
def run(data):
transformers = Network(data, Transformer, " - ", ": ")
if not transformers.is_multi_component():
print("\n".join([v for v in subtask1(transformers)]))
print("\n".join([v for v in subtask2(transformers)]))
else:
print("Error: Input has more than one component")
def test_union(self):
# check toy examples
G1 = Network(vertex_names=["A", "B", "C"],
edges=[("A", "B"), ("B", "C")],
vertex_functions=[sympy.And] * 3)
G2 = Network(vertex_names=["A", "X"],
edges=[("A", "X")],
vertex_functions=[None, sympy.Nor])
sum_true = Network(
vertex_names=["a_A", "a_B", "b_A", "b_X",
"a_C"], # order changed on purpose
edges=[("a_A", "a_B"), ("b_A", "b_X"), ("a_B", "a_C")],
vertex_functions=[
sympy.And, sympy.And, sympy.Nor, sympy.Nor, sympy.And
])
sum_false_1 = Network(vertex_names=["a_A", "a_B", "b_A", "b_X", "a_C"],
edges=[("a_A", "a_B"), ("b_A", "b_X"),
("a_B", "a_C")],
vertex_functions=[
sympy.And, sympy.Nor, sympy.And, sympy.Nor,
sympy.And
])
sum_false_2 = Network(
vertex_names=["a_B", "b_A", "b_X", "a_C"],
edges=[("b_A", "b_X"), ("a_B", "a_C")],
vertex_functions=[sympy.And, sympy.Nor, sympy.Nor, sympy.And])
G_sum = G1 + G2 # order matters, for now.
self.assertEqual(G_sum, sum_true)
self.assertNotEqual(G_sum, sum_false_1)
self.assertNotEqual(G_sum, sum_false_2)
def run(data):
transformers = Network(data, Transformer, " - ", ": ")
if transformers_status_ok(transformers):
print("Stav site OK")
else:
print("Stav site ERROR")
def run(data):
transformers = Network(data, Transformer, DELIM, ": ")
after_reset = deepcopy(transformers) # copy object
after_reset.empty() # empty it
weights, inverted = invert_dict(transformers.vertices)
to_cover = set(transformers.nodes.keys())
covered = set()
while covered != to_cover:
for weight in weights:
for vertex in inverted[weight]:
a, b = vertex.split(DELIM)
new = deepcopy(after_reset)
# connect nodes in graph
new.nodes[a].connect(b)
new.nodes[b].connect(a)
# add vertice to vertices
new.vertices[vertex] = weight
if not new.has_loop():
after_reset = new
covered.add(a)
covered.add(b)
total_cost = 0
for vertex, cost in after_reset.vertices.items():
print(f"{vertex}: {cost}")
total_cost += cost
print("Hodnoceni:", total_cost)
def test_randomize_incoming_edges(self):
for i in range(50):
n = random.randint(1, 10)
G = Network.generate_random(n_vertices=n, indegree_bounds=[1, 3])
G_tag = G.copy()
G_tag.randomize_incoming_edges(include_self_loops=False)
for v, v_tag in zip(G.vertices, G_tag.vertices):
self.assertTrue(v.function == v_tag.function)
self.assertTrue(
len(v.predecessors()) == len(v_tag.predecessors()))
G = Network(vertex_names=["A", "B", "C"],
edges=[("A", "B"), ("A", "A"), ("B", "A")],
vertex_functions=[None] * 3)
for i in range(100):
G_tag = G.copy()
G_tag.randomize_incoming_edges(include_self_loops=False)
self.assertTrue(
G_tag.get_vertex("A").predecessors() ==
[G_tag.get_vertex("B"),
G_tag.get_vertex("C")])
self.assertTrue(
G_tag.get_vertex("B").predecessors() in
[[G_tag.get_vertex("A")], [G_tag.get_vertex("C")]])
self.assertTrue(G_tag.get_vertex("C").predecessors() == [])
had_self_loop = False
for i in range(100):
G_tag = G.copy()
G_tag.randomize_incoming_edges(include_self_loops=True)
self.assertTrue(
G_tag.get_vertex("A").predecessors() in
[[G_tag.get_vertex("A"),
G_tag.get_vertex("B")],
[G_tag.get_vertex("B"),
G_tag.get_vertex("C")],
[G_tag.get_vertex("A"),
G_tag.get_vertex("C")]])
self.assertTrue(
G_tag.get_vertex("B").predecessors() in [[
G_tag.get_vertex("A")
], [G_tag.get_vertex("B")], [G_tag.get_vertex("C")]])
self.assertTrue(G_tag.get_vertex("C").predecessors() == [])
had_self_loop = (G.vertices[0] in G.vertices[0].predecessors()) or \
(G.vertices[1] in G.vertices[1].predecessors()) or had_self_loop
self.assertTrue(had_self_loop)
def test_cnet_export_and_import(self):
for _ in range(10):
n = random.randint(1, 20)
for restriction in [
FunctionTypeRestriction.NONE,
FunctionTypeRestriction.SYMMETRIC_THRESHOLD
]:
# TODO: implement and test for FunctionTypeRestriction.SIMPLE_GATES
if random.choice([False, True]):
indegree_bounds = [0, 4]
indegree_geometric_p = None
else:
indegree_bounds = None
indegree_geometric_p = 0.8
G = Network.generate_random(
n_vertices=n,
indegree_bounds=indegree_bounds,
function_type_restriction=restriction,
indegree_geometric_p=indegree_geometric_p)
G.export_to_cnet("./temp_test_network.cnet")
G_tag = Network.parse_cnet("./temp_test_network.cnet")
self.assertEqual(G, G_tag)
# Network.parse_cnet("C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
# "\\Attractors - for Ariel\\BNS_Dubrova_2011\\MAPK_large2.cnet")
Network.parse_cnet(
"C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
"\\Attractors - for Ariel\\BNS_Dubrova_2011\\arabidopsis.cnet")
Network.parse_cnet(
"C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
"\\Attractors - for Ariel\\BNS_Dubrova_2011\\EGFR_man_rand_inputs1.cnet"
)
# Network.parse_cnet("C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
# "\\Attractors - for Ariel\\BNS_Dubrova_2011\\MAPK_large.cnet")
Network.parse_cnet(
"C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
"\\Attractors - for Ariel\\BNS_Dubrova_2011\\thelper.cnet")
Network.parse_cnet(
"C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
"\\Attractors - for Ariel\\BNS_Dubrova_2011\\tcr.cnet")
def test_truth_table_import_export(self):
for model_dir in os.listdir("../cellcollective_models"):
print model_dir
try:
G = Network.parse_boolean_tables(
os.path.join("../cellcollective_models", model_dir))
except ValueError as e:
if e.message.startswith(
"Model export from cellcollective failed"):
print "\nWarning: {}\n".format(e.message)
pass
else:
raise e
for i in range(50):
n = random.randint(1, 10)
G = Network.generate_random(n_vertices=n, indegree_bounds=[1, 3])
if os.path.exists("temp_dir"):
shutil.rmtree("temp_dir", ignore_errors=True)
G.export_to_boolean_tables(".", "temp_dir")
G_tag = Network.parse_boolean_tables("temp_dir")
self.assertTrue(G == G_tag)
def test_pow_mult(self):
# manual with toy examples
G = Network(vertex_names=["a", "b"],
edges=[("a", "a"), ("a", "b"), ("b", "a")],
vertex_functions=[sympy.And, sympy.And])
G_squared = G * G
# 00 -> 00 -> 00
self.assertTrue(
utility.is_same_state(
G_squared.next_state([False, False], return_as_string=False),
[False, False]))
# 01 -> 00 -> 00
self.assertTrue(
utility.is_same_state(
G_squared.next_state([False, True], return_as_string=False),
[False, False]))
# 10 -> 01 -> 00
self.assertTrue(
utility.is_same_state(
G_squared.next_state([True, False], return_as_string=False),
[False, False]))
# 11 -> 11 -> 11
self.assertTrue(
utility.is_same_state(
G_squared.next_state([True, True], return_as_string=False),
[True, True]))
G = Network(vertex_names=["a", "b"],
edges=[("a", "a"), ("a", "b"), ("b", "a")],
vertex_functions=[sympy.Nand, sympy.And])
G_squared = G * G
# 00 -> 10 -> 11
self.assertTrue(
utility.is_same_state(
G_squared.next_state([False, False], return_as_string=False),
[True, True]))
# 01 -> 10 -> 11
self.assertTrue(
utility.is_same_state(
G_squared.next_state([False, True], return_as_string=False),
[True, True]))
# 10 -> 11 -> 01
self.assertTrue(
utility.is_same_state(
G_squared.next_state([True, False], return_as_string=False),
[False, True]))
# 11 -> 01 -> 10
self.assertTrue(
utility.is_same_state(G_squared.next_state([True, True]),
[True, False]))
# random testing
test_graphs = [Network.parse_cnet("C:\\Users\\ariel\\Downloads\\Attractors - for Ariel"
"\\Attractors - for Ariel\\BNS_Dubrova_2011\\MAPK_large2.cnet")] +\
[Network.generate_random(n_vertices=random.randint(1, 10), indegree_bounds=(0, 5))
for _ in range(5)]
for G in test_graphs:
G_squared = G**2
G_cubed = G**3
for i in range(20):
starting_state = [
bool(random.randint(0, 1)) for _ in range(len(G.vertices))
]
true_double_step = G.next_state(G.next_state(
starting_state, return_as_string=False),
return_as_string=False)
true_triple_step = G.next_state(G.next_state(
G.next_state(starting_state, return_as_string=False),
return_as_string=False),
return_as_string=False)
self.assertTrue(
utility.is_same_state(
G_squared.next_state(starting_state,
return_as_string=False),
true_double_step))
self.assertTrue(
utility.is_same_state(
G_cubed.next_state(starting_state,
return_as_string=False),
true_triple_step))