def test_maxent_1(vars):
"""
Test xor only fixing individual marginals.
"""
d1 = uniform(['000', '011', '101', '110'])
d2 = uniform(['000', '001', '010', '011', '100', '101', '110', '111'])
d1_maxent = maxent_dist(d1, vars)
assert d2.is_approx_equal(d1_maxent, rtol=1e-3, atol=1e-3)
def test_maxent_2():
"""
Text a distribution with differing alphabets.
"""
d1 = uniform(['00', '10', '21', '31'])
d2 = uniform(['00', '01', '10', '11', '20', '21', '30', '31'])
d1_maxent = maxent_dist(d1, [[0], [1]])
assert d2.is_approx_equal(d1_maxent, rtol=1e-3, atol=1e-3)
def test_maxent_2():
"""
Text a distribution with differing alphabets.
"""
d1 = uniform(['00', '10', '21', '31'])
d2 = uniform(['00', '01', '10', '11', '20', '21', '30', '31'])
d1_maxent = maxent_dist(d1, [[0], [1]])
assert d2.is_approx_equal(d1_maxent, rtol=1e-3, atol=1e-3)
def test_maxent_1(vars):
"""
Test xor only fixing individual marginals.
"""
d1 = uniform(['000', '011', '101', '110'])
d2 = uniform(['000', '001', '010', '011', '100', '101', '110', '111'])
d1_maxent = maxent_dist(d1, vars)
assert d2.is_approx_equal(d1_maxent, rtol=1e-3, atol=1e-3)
def test_1():
"""
"""
d = uniform(['0000', '0001', '0010', '0100', '0101', '1000', '1001', '1010'])
assert necessary_conditional_entropy(d, [0,1], [2,3]) == pytest.approx(0.68872187554086695)
assert necessary_conditional_entropy(d, [2,3], [0,1]) == pytest.approx(0.68872187554086695)
assert necessary_conditional_entropy(d, [1], [2,3]) == pytest.approx(0.68872187554086695)
def test_mincoinfo_1():
"""
Test mincoinfo
"""
d = uniform(['000', '111'])
mcio = MinCoInfoOptimizer(d, [[0], [1], [2]])
mcio.optimize()
dp = mcio.construct_dist()
assert I(dp) == pytest.approx(-1)
def test_minent_1():
"""
Test minent
"""
d = uniform(['000', '001', '010', '011', '100', '101', '110', '111'])
meo = MinEntOptimizer(d, [[0], [1], [2]])
meo.optimize()
dp = meo.construct_dist()
assert H(dp) == pytest.approx(1)
def test_dtc_2():
"""
test min dtc
"""
d = uniform(['000', '111'])
max_dtc = MinDualTotalCorrelationOptimizer(d, [[0], [1], [2]])
max_dtc.optimize()
dp = max_dtc.construct_dist()
assert B(dp) == pytest.approx(0.0, abs=1e-4)
def test_dtc_1():
"""
test max dtc
"""
d = uniform(['000', '111'])
max_dtc = MaxDualTotalCorrelationOptimizer(d, [[0], [1], [2]])
max_dtc.optimize()
dp = max_dtc.construct_dist()
assert B(dp) == pytest.approx(2.0, abs=1e-4)
def test_mincoinfo_1():
"""
Test mincoinfo
"""
d = uniform(['000', '111'])
mcio = MinCoInfoOptimizer(d, [[0], [1], [2]])
mcio.optimize()
dp = mcio.construct_dist()
assert I(dp) == pytest.approx(-1)
def test_minent_1():
"""
Test minent
"""
d = uniform(['000', '001', '010', '011', '100', '101', '110', '111'])
meo = MinEntOptimizer(d, [[0], [1], [2]])
meo.optimize()
dp = meo.construct_dist()
assert H(dp) == pytest.approx(1)
def test_maxent_3():
"""
Test the RdnUnqXor distribution.
"""
X00, X01, X02, Y01, Y02 = 'rR', 'aA', [0, 1], 'bB', [0, 1]
inputs = product(X00, X01, X02, Y01, Y02)
events = [(x00+x01+str(x02), x00+y01+str(y02), x00+x01+y01+str(x02^y02)) for x00, x01, x02, y01, y02 in inputs]
RdnUnqXor = uniform(events)
d = maxent_dist(RdnUnqXor, [[0, 1], [0, 2], [1, 2]])
assert H(d) == pytest.approx(6)
def test_maxent_3():
"""
Test the RdnUnqXor distribution.
"""
X00, X01, X02, Y01, Y02 = 'rR', 'aA', [0, 1], 'bB', [0, 1]
inputs = product(X00, X01, X02, Y01, Y02)
events = [(x00 + x01 + str(x02), x00 + y01 + str(y02),
x00 + x01 + y01 + str(x02 ^ y02))
for x00, x01, x02, y01, y02 in inputs]
RdnUnqXor = uniform(events)
d = maxent_dist(RdnUnqXor, [[0, 1], [0, 2], [1, 2]])
assert H(d) == pytest.approx(6)
from hypothesis import given
from dit.distconst import uniform
from dit.multivariate import cohesion, dual_total_correlation, total_correlation
from dit.utils.testing import distributions
reed_solomon = uniform([
'0000',
'0123',
'0231',
'0312',
'1111',
'1032',
'1320',
'1203',
'2222',
'2301',
'2013',
'2130',
'3333',
'3210',
'3102',
'3021',
])
@given(dist=distributions(alphabets=(2, ) * 4))
def test_cohesion_1(dist):
"""
Test that k=1 is the total correlation.
"""
