def test_clear_subsystem_caches_after_computing_sia_config_option(s):
with config.override(CLEAR_SUBSYSTEM_CACHES_AFTER_COMPUTING_SIA=False,
PARALLEL_CONCEPT_EVALUATION=False,
PARALLEL_CUT_EVALUATION=False,
CACHE_REPERTOIRES=True):
sia = compute.sia(s)
assert s._repertoire_cache.cache
with config.override(CLEAR_SUBSYSTEM_CACHES_AFTER_COMPUTING_SIA=True,
PARALLEL_CONCEPT_EVALUATION=False,
PARALLEL_CUT_EVALUATION=False,
CACHE_REPERTOIRES=True):
sia = compute.sia(s)
assert not s._repertoire_cache.cache
def test_sia_disconnected_network(reducible):
assert (compute.sia(reducible) ==
models.SystemIrreducibilityAnalysis(subsystem=reducible,
cut_subsystem=reducible,
phi=0.0,
ces=[],
partitioned_ces=[]))
def test_sia_disconnected_network(reducible):
assert compute.sia(reducible) == models.SystemIrreducibilityAnalysis(
subsystem=reducible,
cut_subsystem=reducible,
phi=0.0,
ces=[],
partitioned_ces=[],
)
def test_sia_empty_subsystem(s_empty):
assert (compute.sia(s_empty) ==
models.SystemIrreducibilityAnalysis(
phi=0.0,
ces=(),
partitioned_ces=(),
subsystem=s_empty,
cut_subsystem=s_empty))
def test_sia_wrappers(reducible):
assert (compute.sia(reducible) ==
models.SystemIrreducibilityAnalysis(subsystem=reducible,
cut_subsystem=reducible,
phi=0.0,
ces=[],
partitioned_ces=[]))
assert compute.phi(reducible) == 0.0
def test_degenerate(degenerate):
assert np.array_equal(
degenerate.tpm,
convert.to_multidimensional(np.array([[0, 0], [0, 1], [1, 0], [1,
1]])))
assert np.array_equal(degenerate.cm, np.array([[0, 1], [1, 0]]))
sia = compute.sia(degenerate)
assert sia.phi == 0.638888
def test_sia_disconnected_network(reducible, flushcache, restore_fs_cache):
flushcache()
assert (compute.sia(reducible) == models.SystemIrreducibilityAnalysis(
subsystem=reducible,
cut_subsystem=reducible,
phi=0.0,
ces=[],
partitioned_ces=[]))
def test_sia_empty_subsystem(s_empty, flushcache, restore_fs_cache):
flushcache()
assert (compute.sia(s_empty) == models.SystemIrreducibilityAnalysis(
phi=0.0,
ces=(),
partitioned_ces=(),
subsystem=s_empty,
cut_subsystem=s_empty))
def test_sia_wrappers(reducible):
assert (compute.sia(reducible) ==
models.SystemIrreducibilityAnalysis(subsystem=reducible,
cut_subsystem=reducible,
phi=0.0,
ces=[],
partitioned_ces=[]))
assert compute.phi(reducible) == 0.0
def test_sia_empty_subsystem(s_empty):
assert (compute.sia(s_empty) ==
models.SystemIrreducibilityAnalysis(
phi=0.0,
ces=(),
partitioned_ces=(),
subsystem=s_empty,
cut_subsystem=s_empty))
def test_ces_distance_switches_to_small_phi_difference(s):
sia = compute.sia(s)
ce_structures = (sia.ces, sia.partitioned_ces)
with config.override(USE_SMALL_PHI_DIFFERENCE_FOR_CES_DISTANCE=False):
assert compute.ces_distance(*ce_structures) == 2.3125
with config.override(USE_SMALL_PHI_DIFFERENCE_FOR_CES_DISTANCE=True):
assert compute.ces_distance(*ce_structures) == 1.083333
def test_clear_subsystem_caches_after_computing_sia_config_option(s):
with config.override(
CLEAR_SUBSYSTEM_CACHES_AFTER_COMPUTING_SIA=False,
PARALLEL_CONCEPT_EVALUATION=False,
PARALLEL_CUT_EVALUATION=False,
CACHE_REPERTOIRES=True,
):
sia = compute.sia(s)
assert s._repertoire_cache.cache
with config.override(
CLEAR_SUBSYSTEM_CACHES_AFTER_COMPUTING_SIA=True,
PARALLEL_CONCEPT_EVALUATION=False,
PARALLEL_CUT_EVALUATION=False,
CACHE_REPERTOIRES=True,
):
sia = compute.sia(s)
assert not s._repertoire_cache.cache
def test_xor_propogation_delay():
# Three interconnected XOR gates, with COPY gates along each connection
# acting as propagation delays.
nodes = 9
tpm = np.zeros((2 ** nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[2] ^ ps[7]:
cs[0] = 1
if ps[0] == 1:
cs[1] = 1
cs[8] = 1
if ps[1] ^ ps[5]:
cs[3] = 1
if ps[3] == 1:
cs[2] = 1
cs[4] = 1
if ps[4] ^ ps[8]:
cs[6] = 1
if ps[6] == 1:
cs[5] = 1
cs[7] = 1
tpm[psi, :] = cs
cm = np.array([
[0, 1, 0, 0, 0, 0, 0, 0, 1],
[0, 0, 0, 1, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0],
[0, 0, 0, 1, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 1, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0]
])
# The state of the system is all OFF
state = (0, 0, 0, 0, 0, 0, 0, 0, 0)
network = Network(tpm, cm=cm)
partition = ((0, 2, 7), (1, 3, 5), (4, 6, 8))
output_indices = (0, 3, 6)
blackbox = macro.Blackbox(partition, output_indices)
assert blackbox.hidden_indices == (1, 2, 4, 5, 7, 8)
time = 2
subsys = macro.MacroSubsystem(network, state, network.node_indices,
blackbox=blackbox, time_scale=time)
sia = compute.sia(subsys)
assert sia.phi == 1.874999
assert sia.cut == models.Cut((0,), (1, 2, 3, 4, 5, 6, 7, 8))
def test_deserialization_memoizes_duplicate_objects(s):
with config.override(PARALLEL_CUT_EVALUATION=True):
sia = compute.sia(s)
loaded = jsonify.loads(jsonify.dumps(sia))
l1 = loaded.subsystem
l2 = loaded.ces.subsystem
assert l1 == l2
assert hash(l1) == hash(l2)
assert l1 is l2
def test_xor_propogation_delay():
# Three interconnected XOR gates, with COPY gates along each connection
# acting as propagation delays.
nodes = 9
tpm = np.zeros((2**nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[2] ^ ps[7]:
cs[0] = 1
if ps[0] == 1:
cs[1] = 1
cs[8] = 1
if ps[1] ^ ps[5]:
cs[3] = 1
if ps[3] == 1:
cs[2] = 1
cs[4] = 1
if ps[4] ^ ps[8]:
cs[6] = 1
if ps[6] == 1:
cs[5] = 1
cs[7] = 1
tpm[psi, :] = cs
cm = np.array([[0, 1, 0, 0, 0, 0, 0, 0, 1], [0, 0, 0, 1, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 1, 0, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0], [0, 0, 0, 1, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 1, 0], [1, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0]])
# The state of the system is all OFF
state = (0, 0, 0, 0, 0, 0, 0, 0, 0)
network = Network(tpm, cm=cm)
partition = ((0, 2, 7), (1, 3, 5), (4, 6, 8))
output_indices = (0, 3, 6)
blackbox = macro.Blackbox(partition, output_indices)
assert blackbox.hidden_indices == (1, 2, 4, 5, 7, 8)
time = 2
subsys = macro.MacroSubsystem(network,
state,
network.node_indices,
blackbox=blackbox,
time_scale=time)
sia = compute.sia(subsys)
assert sia.phi == 1.874999
assert sia.cut == models.Cut((0, ), (1, 2, 3, 4, 5, 6, 7, 8))
def test_degenerate(degenerate):
assert np.array_equal(degenerate.tpm, convert.to_multidimensional(np.array([
[0, 0],
[0, 1],
[1, 0],
[1, 1]
])))
assert np.array_equal(degenerate.cm, np.array([
[0, 1],
[1, 0]
]))
sia = compute.sia(degenerate)
assert sia.phi == 0.638888
def test_expand_cause_repertoire(micro_s_all_off):
sia = compute.sia(micro_s_all_off)
A = sia.ces[0]
cause = A.cause_repertoire
assert np.all(abs(A.expand_cause_repertoire(CD) - cause) < EPSILON)
assert np.all(abs(
A.expand_cause_repertoire(BCD).flatten(order='F') -
np.array([1 / 6 if i < 6 else 0 for i in range(8)])) < EPSILON)
assert np.all(abs(
A.expand_cause_repertoire(ABCD).flatten(order='F') -
np.array([1 / 12 if i < 12 else 0 for i in range(16)])) < EPSILON)
assert np.all(abs(A.expand_cause_repertoire(ABCD) -
A.expand_cause_repertoire()) < EPSILON)
def test_coarsegrain_spatial_degenerate():
# TODO: move to docs?
# macro-micro examples from Hoel2016
# Example 2 - Spatial Degenerate
# The micro system has a full complex, and big_phi = 0.19
# The optimal coarse-graining groups AB, CD and EF, each with state
# mapping ((0, 1), (2))
nodes = 6
tpm = np.zeros((2**nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[0] == 1 and ps[1] == 1:
cs[2] = 1
cs[3] = 1
if ps[2] == 1 and ps[3] == 1:
cs[4] = 1
cs[5] = 1
if ps[4] == 1 and ps[5] == 1:
cs[0] = 1
cs[1] = 1
tpm[psi, :] = cs
cm = np.array([
[0, 0, 1, 1, 0, 0],
[0, 0, 1, 1, 0, 0],
[0, 0, 0, 0, 1, 1],
[0, 0, 0, 0, 1, 1],
[1, 1, 0, 0, 0, 0],
[1, 1, 0, 0, 0, 0]
])
state = (0, 0, 0, 0, 0, 0)
net = Network(tpm, cm)
mc = compute.major_complex(net, state)
assert mc.phi == 0.194445
partition = ((0, 1), (2, 3), (4, 5))
grouping = (((0, 1), (2,)), ((0, 1), (2,)), ((0, 1), (2, )))
coarse = macro.CoarseGrain(partition, grouping)
sub = macro.MacroSubsystem(net, state, range(net.size),
coarse_grain=coarse)
sia = compute.sia(sub)
assert sia.phi == 0.834183
def test_expand_cause_repertoire(micro_s_all_off):
sia = compute.sia(micro_s_all_off)
A = sia.ces[0]
cause = A.cause_repertoire
assert np.all(abs(A.expand_cause_repertoire(CD) - cause) < EPSILON)
assert np.all(
abs(
A.expand_cause_repertoire(BCD).flatten(order='F') -
np.array([1 / 6 if i < 6 else 0 for i in range(8)])) < EPSILON)
assert np.all(
abs(
A.expand_cause_repertoire(ABCD).flatten(order='F') -
np.array([1 / 12 if i < 12 else 0 for i in range(16)])) < EPSILON)
assert np.all(
abs(A.expand_cause_repertoire(ABCD) -
A.expand_cause_repertoire()) < EPSILON)
def test_expand_effect_repertoire(micro_s_all_off):
sia = compute.sia(micro_s_all_off)
A = sia.ces[0]
effect = A.effect_repertoire
assert np.all(abs(A.expand_effect_repertoire(CD) - effect) < EPSILON)
assert np.all(abs(A.expand_effect_repertoire(BCD).flatten(order='F') -
np.array([.25725, .23275, .11025, .09975,
.11025, .09975, .04725, .04275])) < EPSILON)
assert np.all(abs(
A.expand_effect_repertoire(ABCD).flatten(order='F') -
np.array([.13505625, .12219375, .12219375, .11055625,
.05788125, .05236875, .05236875, .04738125,
.05788125, .05236875, .05236875, .04738125,
.02480625, .02244375, .02244375, .02030625])) < EPSILON)
assert np.all(abs(A.expand_effect_repertoire(ABCD) -
A.expand_effect_repertoire()) < EPSILON)
def test_coarsegrain_spatial_degenerate():
# TODO: move to docs?
# macro-micro examples from Hoel2016
# Example 2 - Spatial Degenerate
# The micro system has a full complex, and big_phi = 0.19
# The optimal coarse-graining groups AB, CD and EF, each with state
# mapping ((0, 1), (2))
nodes = 6
tpm = np.zeros((2**nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[0] == 1 and ps[1] == 1:
cs[2] = 1
cs[3] = 1
if ps[2] == 1 and ps[3] == 1:
cs[4] = 1
cs[5] = 1
if ps[4] == 1 and ps[5] == 1:
cs[0] = 1
cs[1] = 1
tpm[psi, :] = cs
cm = np.array([[0, 0, 1, 1, 0, 0], [0, 0, 1, 1, 0, 0], [0, 0, 0, 0, 1, 1],
[0, 0, 0, 0, 1, 1], [1, 1, 0, 0, 0, 0], [1, 1, 0, 0, 0, 0]])
state = (0, 0, 0, 0, 0, 0)
net = Network(tpm, cm)
mc = compute.major_complex(net, state)
assert mc.phi == 0.194445
partition = ((0, 1), (2, 3), (4, 5))
grouping = (((0, 1), (2, )), ((0, 1), (2, )), ((0, 1), (2, )))
coarse = macro.CoarseGrain(partition, grouping)
sub = macro.MacroSubsystem(net,
state,
range(net.size),
coarse_grain=coarse)
sia = compute.sia(sub)
assert sia.phi == 0.834183
def test_deserialization_memoizes_duplicate_objects(s):
with config.override(PARALLEL_CUT_EVALUATION=True):
sia = compute.sia(s)
s1 = sia.subsystem
# Computed in a parallel process, so has a different id
s2 = sia.ces[0].subsystem
assert s1 is not s2
assert s1 == s2
assert hash(s1) == hash(s2)
loaded = jsonify.loads(jsonify.dumps(sia))
l1 = loaded.subsystem
l2 = loaded.ces[0].subsystem
assert l1 == l2
assert hash(l1) == hash(l2)
assert l1 is l2
def test_expand_effect_repertoire(micro_s_all_off):
sia = compute.sia(micro_s_all_off)
A = sia.ces[0]
effect = A.effect_repertoire
assert np.all(abs(A.expand_effect_repertoire(CD) - effect) < EPSILON)
assert np.all(
abs(
A.expand_effect_repertoire(BCD).flatten(order='F') - np.array([
.25725, .23275, .11025, .09975, .11025, .09975, .04725, .04275
])) < EPSILON)
assert np.all(
abs(
A.expand_effect_repertoire(ABCD).flatten(order='F') - np.array([
.13505625, .12219375, .12219375, .11055625, .05788125,
.05236875, .05236875, .04738125, .05788125, .05236875,
.05236875, .04738125, .02480625, .02244375, .02244375,
.02030625
])) < EPSILON)
assert np.all(
abs(A.expand_effect_repertoire(ABCD) -
A.expand_effect_repertoire()) < EPSILON)
def test_sia_macro(macro_s):
sia = compute.sia(macro_s)
check_sia(sia, macro_answer)
def test_sia_rule152(rule152_s):
sia = compute.sia(rule152_s)
check_sia(sia, rule152_answer)
def test_sia_big_network_0_thru_3(big_subsys_0_thru_3):
sia = compute.sia(big_subsys_0_thru_3)
check_sia(sia, big_subsys_0_thru_3_answer)
def test_sia_big_network(big_subsys_all):
sia = compute.sia(big_subsys_all)
check_sia(sia, big_answer)
def test_sia_complete_graph_rule152_s(rule152_s_complete):
sia = compute.sia(rule152_s_complete)
check_sia(sia, rule152_answer)
def test_basic_nor_or():
# A system composed of NOR and OR (copy) gates, which mimics the basic
# pyphi network
nodes = 12
tpm = np.zeros((2**nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[5] == 0 and ps[11] == 0:
cs[0] = 1
if ps[0] == 0:
cs[1] = 1
if ps[1] == 1:
cs[2] = 1
if ps[11] == 0:
cs[3] = 1
if ps[3] == 0:
cs[4] = 1
if ps[4] == 1:
cs[5] = 1
if ps[2] == 0:
cs[6] = 1
if ps[5] == 0:
cs[7] = 1
if ps[6] == 0 and ps[7] == 0:
cs[8] = 1
if ps[2] == 0 and ps[5] == 0:
cs[9] = 1
if ps[9] == 1:
cs[10] = 1
if ps[8] == 0 and ps[10] == 0:
cs[11] = 1
tpm[psi, :] = cs
cm = np.array([
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0],
[0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0],
])
state = (0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0)
network = Network(tpm, cm=cm)
# (0, 1, 2) compose the OR element,
# (3, 4, 5) the COPY,
# (6, 7, 8, 9, 10, 11) the XOR
partition = ((0, 1, 2), (3, 4, 5), (6, 7, 8, 9, 10, 11))
output = (2, 5, 11)
blackbox = macro.Blackbox(partition, output)
assert blackbox.hidden_indices == (0, 1, 3, 4, 6, 7, 8, 9, 10)
time = 3
sub = macro.MacroSubsystem(network,
state,
network.node_indices,
blackbox=blackbox,
time_scale=time)
with config.override(CUT_ONE_APPROXIMATION=True):
sia = compute.sia(sub)
assert sia.phi == 1.958332
assert sia.cut == models.Cut((6, ), (0, 1, 2, 3, 4, 5, 7, 8, 9, 10, 11))
def test_sia_complete_graph_standard_example(s_complete):
sia = compute.sia(s_complete)
check_sia(sia, standard_answer)
def test_sia_single_micro_nodes_without_selfloops_dont_have_phi(s_single):
assert compute.sia(s_single).phi == 0.0
def test_sia_single_micro_node_selfloops_dont_have_phi(noisy_selfloop_single):
assert compute.sia(noisy_selfloop_single).phi == 0.0
def test_sia_big_network_0_thru_3(big_subsys_0_thru_3, flushcache,
restore_fs_cache):
flushcache()
sia = compute.sia(big_subsys_0_thru_3)
check_sia(sia, big_subsys_0_thru_3_answer)
def test_sia_rule152(rule152_s, flushcache, restore_fs_cache):
flushcache()
sia = compute.sia(rule152_s)
check_sia(sia, rule152_answer)
def test_sia_macro(macro_s, flushcache, restore_fs_cache):
flushcache()
sia = compute.sia(macro_s)
check_sia(sia, macro_answer)
def test_basic_nor_or():
# A system composed of NOR and OR (copy) gates, which mimics the basic
# pyphi network
nodes = 12
tpm = np.zeros((2 ** nodes, nodes))
for psi, ps in enumerate(utils.all_states(nodes)):
cs = [0 for i in range(nodes)]
if ps[5] == 0 and ps[11] == 0:
cs[0] = 1
if ps[0] == 0:
cs[1] = 1
if ps[1] == 1:
cs[2] = 1
if ps[11] == 0:
cs[3] = 1
if ps[3] == 0:
cs[4] = 1
if ps[4] == 1:
cs[5] = 1
if ps[2] == 0:
cs[6] = 1
if ps[5] == 0:
cs[7] = 1
if ps[6] == 0 and ps[7] == 0:
cs[8] = 1
if ps[2] == 0 and ps[5] == 0:
cs[9] = 1
if ps[9] == 1:
cs[10] = 1
if ps[8] == 0 and ps[10] == 0:
cs[11] = 1
tpm[psi, :] = cs
cm = np.array([
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0],
[0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0],
[1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
[1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0],
])
state = (0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0)
network = Network(tpm, cm=cm)
# (0, 1, 2) compose the OR element,
# (3, 4, 5) the COPY,
# (6, 7, 8, 9, 10, 11) the XOR
partition = ((0, 1, 2), (3, 4, 5), (6, 7, 8, 9, 10, 11))
output = (2, 5, 11)
blackbox = macro.Blackbox(partition, output)
assert blackbox.hidden_indices == (0, 1, 3, 4, 6, 7, 8, 9, 10)
time = 3
sub = macro.MacroSubsystem(network, state, network.node_indices,
blackbox=blackbox, time_scale=time)
with config.override(CUT_ONE_APPROXIMATION=True):
sia = compute.sia(sub)
assert sia.phi == 1.958332
assert sia.cut == models.Cut((6,), (0, 1, 2, 3, 4, 5, 7, 8, 9, 10, 11))
def test_unpickle(s):
bm = compute.sia(s)
pickle.loads(pickle.dumps(bm))
def test_sia_big_network(big_subsys_all, flushcache, restore_fs_cache):
flushcache()
sia = compute.sia(big_subsys_all)
check_sia(sia, big_answer)
def test_sia_complete_graph_s_noised(s_noised_complete):
sia = compute.sia(s_noised_complete)
check_sia(sia, noised_answer)
def test_unpickle(s):
bm = compute.sia(s)
pickle.loads(pickle.dumps(bm))
def test_sia_complete_graph_big_subsys_all(big_subsys_all_complete):
sia = compute.sia(big_subsys_all_complete)
check_sia(sia, big_answer)
