本文整理汇总了Python中bayespy.nodes.Mixture类的典型用法代码示例。如果您正苦于以下问题：Python Mixture类的具体用法？Python Mixture怎么用？Python Mixture使用的例子？那么恭喜您, 这里精选的类代码示例或许可以为您提供帮助。
在下文中一共展示了Mixture类的15个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: test_mask_to_parent
def test_mask_to_parent(self):
"""
Test the mask handling in Mixture node
"""
K = 3
Z = Categorical(np.ones(K)/K,
plates=(4,5))
Mu = GaussianARD(0, 1,
shape=(2,),
plates=(4,K,5))
Alpha = Gamma(1, 1,
plates=(4,K,5,2))
X = Mixture(Z, GaussianARD, Mu, Alpha, cluster_plate=-2)
Y = GaussianARD(X, 1)
mask = np.reshape((np.mod(np.arange(4*5), 2) == 0),
(4,5))
Y.observe(np.ones((4,5,2)),
mask=mask)
self.assertArrayEqual(X._mask_to_parent(0),
mask)
self.assertArrayEqual(X._mask_to_parent(1),
mask[:,None,:])
self.assertArrayEqual(X._mask_to_parent(2),
mask[:,None,:,None])
pass示例2: test_mixture
def test_mixture(self):
"""
Test mixture of Bernoulli
"""
P = Mixture([2,0,0], Bernoulli, [0.1, 0.2, 0.3])
u = P._message_to_child()
self.assertEqual(len(u), 1)
self.assertAllClose(u[0], [0.3, 0.1, 0.1])
pass示例3: test_mixture
def test_mixture(self):
"""
Test binomial mixture
"""
X = Mixture(2, Binomial, 10, [0.1, 0.2, 0.3, 0.4])
u = X._message_to_child()
self.assertAllClose(u[0], 3.0)
pass示例4: test_lowerbound
def test_lowerbound(self):
"""
Test log likelihood lower bound for Mixture node
"""
# Mixed distribution broadcasts g
# This tests for a found bug. The bug caused an error.
Z = Categorical([0.3, 0.5, 0.2])
X = Mixture(Z, Categorical, [[0.2,0.8], [0.1,0.9], [0.3,0.7]])
X.lower_bound_contribution()
pass示例5: test_random
def test_random(self):
"""
Test random sampling of mixture node
"""
o = 1e-20
X = Mixture([1, 0, 2], Categorical, [ [o, o, o, 1],
[o, o, 1, o],
[1, o, o, o] ])
x = X.random()
self.assertAllClose(x, [2, 3, 0])
pass示例6: test_mixture
def test_mixture(self):
"""
Test multinomial mixture
"""
p0 = [0.1, 0.5, 0.2, 0.2]
p1 = [0.5, 0.1, 0.1, 0.3]
p2 = [0.3, 0.2, 0.1, 0.4]
X = Mixture(2, Multinomial, 10, [p0, p1, p2])
u = X._message_to_child()
self.assertAllClose(u[0],
10*np.array(p2))
pass示例7: test_message_to_child
def test_message_to_child(self):
"""
Test the message to child of Mixture node.
"""
K = 3
#
# Estimate moments from parents only
#
# Simple case
mu = GaussianARD([0,2,4], 1,
ndim=0,
plates=(K,))
alpha = Gamma(1, 1,
plates=(K,))
z = Categorical(np.ones(K)/K)
X = Mixture(z, GaussianARD, mu, alpha)
self.assertEqual(X.plates, ())
self.assertEqual(X.dims, ( (), () ))
u = X._message_to_child()
self.assertAllClose(u[0],
2)
self.assertAllClose(u[1],
2**2+1)
# Broadcasting the moments on the cluster axis
mu = GaussianARD(2, 1,
ndim=0,
plates=(K,))
alpha = Gamma(1, 1,
plates=(K,))
z = Categorical(np.ones(K)/K)
X = Mixture(z, GaussianARD, mu, alpha)
self.assertEqual(X.plates, ())
self.assertEqual(X.dims, ( (), () ))
u = X._message_to_child()
self.assertAllClose(u[0],
2)
self.assertAllClose(u[1],
2**2+1)
#
# Estimate moments with observed children
#
pass示例8: test_gaussian_mixture_plot
def test_gaussian_mixture_plot():
"""
Test the gaussian_mixture plotting function.
The code is from http://www.bayespy.org/examples/gmm.html
"""
np.random.seed(1)
y0 = np.random.multivariate_normal([0, 0], [[1, 0], [0, 0.02]], size=50)
y1 = np.random.multivariate_normal([0, 0], [[0.02, 0], [0, 1]], size=50)
y2 = np.random.multivariate_normal([2, 2], [[1, -0.9], [-0.9, 1]], size=50)
y3 = np.random.multivariate_normal([-2, -2], [[0.1, 0], [0, 0.1]], size=50)
y = np.vstack([y0, y1, y2, y3])
bpplt.pyplot.plot(y[:,0], y[:,1], 'rx')
N = 200
D = 2
K = 10
alpha = Dirichlet(1e-5*np.ones(K),
name='alpha')
Z = Categorical(alpha,
plates=(N,),
name='z')
mu = Gaussian(np.zeros(D), 1e-5*np.identity(D),
plates=(K,),
name='mu')
Lambda = Wishart(D, 1e-5*np.identity(D),
plates=(K,),
name='Lambda')
Y = Mixture(Z, Gaussian, mu, Lambda,
name='Y')
Z.initialize_from_random()
Q = VB(Y, mu, Lambda, Z, alpha)
Y.observe(y)
Q.update(repeat=1000)
bpplt.gaussian_mixture_2d(Y, scale=2)
# Have to define these limits because on some particular environments these
# may otherwise differ and thus result in an image comparsion failure
bpplt.pyplot.xlim([-3, 6])
bpplt.pyplot.ylim([-3, 5])示例9: test_deterministic_mappings
def test_deterministic_mappings(self):
x = Categorical([0.8, 0.2])
y = Mixture(
x,
Categorical,
[
[0.10, 0.90],
[0.00, 1.00],
]
)
y.observe(0)
x.update()
self.assertAllClose(x.u[0], [1, 0])
y.observe(1)
x.update()
p = np.array([0.8*0.9, 0.2*1.0])
self.assertAllClose(x.u[0], p / np.sum(p))
pass示例10: test_nans
def test_nans(self):
"""
Test multinomial mixture
"""
# The probabilities p1 cause problems
p0 = [0.1, 0.9]
p1 = [1.0-1e-50, 1e-50]
Z = Categorical([1-1e-10, 1e-10])
X = Mixture(Z, Multinomial, 10, [p0, p1])
u = X._message_to_child()
self.assertAllClose(u[0],
[1, 9])
p0 = [0.1, 0.9]
p1 = [1.0-1e-10, 1e-10]
Z = Categorical([1-1e-50, 1e-50])
X = Mixture(Z, Multinomial, 10, [p0, p1])
u = X._message_to_child()
self.assertAllClose(u[0],
[1, 9])
with warnings.catch_warnings():
warnings.simplefilter("ignore", RuntimeWarning)
warnings.simplefilter("ignore", UserWarning)
p0 = [0.1, 0.9]
p1 = [1.0, 0.0]
X = Mixture(0, Multinomial, 10, [p0, p1])
u = X._message_to_child()
self.assertAllClose(u[0],
np.nan*np.ones(2))
pass示例11: test_constant
def test_constant(self):
"""
Test constant categorical nodes
"""
# Basic test
Y = Mixture(2, Gamma, [1, 2, 3], [1, 1, 1])
u = Y._message_to_child()
self.assertAllClose(u[0],
3/1)
# Test with one plate axis
alpha = [[1, 2, 3],
[4, 5, 6]]
Y = Mixture([2, 1], Gamma, alpha, 1)
u = Y._message_to_child()
self.assertAllClose(u[0],
[3, 5])
# Test with two plate axes
alpha = [ [[1, 2, 3],
[4, 5, 6]],
[[7, 8, 9],
[10, 11, 12]] ]
Y = Mixture([[2, 1], [0, 2]], Gamma, alpha, 1)
u = Y._message_to_child()
self.assertAllClose(u[0],
[[3, 5],
[7, 12]])
pass示例12: test_gradient
def test_gradient(self):
"""
Check the Euclidean gradient of the categorical node
"""
Z = Categorical([[0.3, 0.5, 0.2], [0.1, 0.6, 0.3]])
Y = Mixture(Z, Gamma, [2, 3, 4], [5, 6, 7])
Y.observe([4.2, 0.2])
def f(x):
Z.set_parameters([np.reshape(x, Z.get_shape(0))])
return Z.lower_bound_contribution() + Y.lower_bound_contribution()
def df(x):
Z.set_parameters([np.reshape(x, Z.get_shape(0))])
g = Z.get_riemannian_gradient()
return Z.get_gradient(g)[0]
x0 = np.ravel(np.log([[2, 3, 7], [0.1, 3, 1]]))
self.assertAllClose(
misc.gradient(f, x0),
np.ravel(df(x0))
)
pass示例13: _setup_bernoulli_mixture
def _setup_bernoulli_mixture():
"""
Setup code for the hinton tests.
This code is from http://www.bayespy.org/examples/bmm.html
"""
np.random.seed(1)
p0 = [0.1, 0.9, 0.1, 0.9, 0.1, 0.9, 0.1, 0.9, 0.1, 0.9]
p1 = [0.1, 0.1, 0.1, 0.1, 0.1, 0.9, 0.9, 0.9, 0.9, 0.9]
p2 = [0.9, 0.9, 0.9, 0.9, 0.9, 0.1, 0.1, 0.1, 0.1, 0.1]
p = np.array([p0, p1, p2])
z = random.categorical([1/3, 1/3, 1/3], size=100)
x = random.bernoulli(p[z])
N = 100
D = 10
K = 10
R = Dirichlet(K*[1e-5],
name='R')
Z = Categorical(R,
plates=(N,1),
name='Z')
P = Beta([0.5, 0.5],
plates=(D,K),
name='P')
X = Mixture(Z, Bernoulli, P)
Q = VB(Z, R, X, P)
P.initialize_from_random()
X.observe(x)
Q.update(repeat=1000)
return (R,P,Z)示例14: test_message_to_parent
def test_message_to_parent(self):
"""
Test the message to parents of Mixture node.
"""
K = 3
# Broadcasting the moments on the cluster axis
Mu = GaussianARD(2, 1,
ndim=0,
plates=(K,))
(mu, mumu) = Mu._message_to_child()
Alpha = Gamma(3, 1,
plates=(K,))
(alpha, logalpha) = Alpha._message_to_child()
z = Categorical(np.ones(K)/K)
X = Mixture(z, GaussianARD, Mu, Alpha)
tau = 4
Y = GaussianARD(X, tau)
y = 5
Y.observe(y)
(x, xx) = X._message_to_child()
m = z._message_from_children()
self.assertAllClose(m[0] * np.ones(K),
random.gaussian_logpdf(xx*alpha,
x*alpha*mu,
mumu*alpha,
logalpha,
0)
* np.ones(K))
m = Mu._message_from_children()
self.assertAllClose(m[0],
1/K * (alpha*x) * np.ones(3))
self.assertAllClose(m[1],
-0.5 * 1/K * alpha * np.ones(3))
# Some parameters do not have cluster plate axis
Mu = GaussianARD(2, 1,
ndim=0,
plates=(K,))
(mu, mumu) = Mu._message_to_child()
Alpha = Gamma(3, 1) # Note: no cluster plate axis!
(alpha, logalpha) = Alpha._message_to_child()
z = Categorical(np.ones(K)/K)
X = Mixture(z, GaussianARD, Mu, Alpha)
tau = 4
Y = GaussianARD(X, tau)
y = 5
Y.observe(y)
(x, xx) = X._message_to_child()
m = z._message_from_children()
self.assertAllClose(m[0] * np.ones(K),
random.gaussian_logpdf(xx*alpha,
x*alpha*mu,
mumu*alpha,
logalpha,
0)
* np.ones(K))
m = Mu._message_from_children()
self.assertAllClose(m[0],
1/K * (alpha*x) * np.ones(3))
self.assertAllClose(m[1],
-0.5 * 1/K * alpha * np.ones(3))
# Cluster assignments do not have as many plate axes as parameters.
M = 2
Mu = GaussianARD(2, 1,
ndim=0,
plates=(K,M))
(mu, mumu) = Mu._message_to_child()
Alpha = Gamma(3, 1,
plates=(K,M))
(alpha, logalpha) = Alpha._message_to_child()
z = Categorical(np.ones(K)/K)
X = Mixture(z, GaussianARD, Mu, Alpha, cluster_plate=-2)
tau = 4
Y = GaussianARD(X, tau)
y = 5 * np.ones(M)
Y.observe(y)
(x, xx) = X._message_to_child()
m = z._message_from_children()
self.assertAllClose(m[0]*np.ones(K),
np.sum(random.gaussian_logpdf(xx*alpha,
x*alpha*mu,
mumu*alpha,
logalpha,
0) *
np.ones((K,M)),
axis=-1))
m = Mu._message_from_children()
self.assertAllClose(m[0] * np.ones((K,M)),
1/K * (alpha*x) * np.ones((K,M)))
self.assertAllClose(m[1] * np.ones((K,M)),
-0.5 * 1/K * alpha * np.ones((K,M)))
# Mixed distribution broadcasts g
# This tests for a found bug. The bug caused an error.
#.........这里部分代码省略.........
示例15: generateCPD
def generateCPD(self, term): #, X_train, y_train, X_test, y_test, X_validation, y_validation, g_train, g_test, g_validation):
clf = loadClf(self.ontology[term]['name'], self.fold, self.clfName)
posTrain = sum(clf.y_train == POSTIVE_LABEL)
negTrain = sum(clf.y_train == NEGATIVE_LABEL)
totalTrain = posTrain + negTrain
children = sorted(self.ontology[term]['children'])
parents = sorted(self.ontology[term]['parents'])
labels = {l : PRIOR for l in product(*((POSTIVE_LABEL,NEGATIVE_LABEL),)*(len(children)+1))}
if children:
childNodes = [self.ontology[child]['node'][self.fold][self.clfName] for child in children]
for gene,y in zip(clf.g_train, clf.y_train):
event = []
for child in children:
event.append(POSTIVE_LABEL if gene in self.ontology.associations[child] else NEGATIVE_LABEL)
event.append(POSTIVE_LABEL if gene in self.ontology.associations[term] else NEGATIVE_LABEL)
assert (gene in self.ontology.associations[term]) == (y == POSTIVE_LABEL)
event = tuple(event)
labels[event] += 1
def countBoth(event):
return labels[event[:-1]+(POSTIVE_LABEL,)] + labels[event[:-1]+(NEGATIVE_LABEL,)]
cprior = PRIOR * (2 ** len(children))
types = [Mixture]*(len(children)-1) + [Categorical]
mixparams = [i for s in zip(childNodes, types) for i in s]
cpd = numpy.empty((2,)*(len(children)+1))
for event, counted in labels.items():
v=cpd
for b in event[:-1]:
v = v[b]
hid = event[-1]
print("Event: ", event)
if POSTIVE_LABEL not in event[:-1]: # Všichni potomci označeni "ne"
v[hid] = counted/countBoth(event)
print("Stored %d / %d" % (counted,countBoth(event)))
else:
v[hid] = {POSTIVE_LABEL: 0.99, NEGATIVE_LABEL:0.01}[hid]
print("Stored %d : %d" % (hid, v[hid]))
#print(term,"<-",",".join(children))
print(cpd)
#print(labels)
hidden = Mixture(*mixparams, cpd)
hidden.params = cpd
else: #No children
#hidden = DiscreteDistribution({'0': posTrain / totalTrain, '1': negTrain / totalTrain})
params = (posTrain / totalTrain, negTrain / totalTrain)
hidden = Categorical(params)
hidden.params = params
#print("Hidden node %s:" % term)
#print(repr(hidden))
#print([p for p in hidden.parents if isinstance(p, Stochastic)])
#print(hidden.get_moments())
conf = clf.conf + PRIOR
#posTest, negTest = numpy.sum(conf, 1)
posTest, negTest = numpy.sum(conf, 0)
#print("Confusion matrix:")
#print(conf)
try:
assert term != self.ontology.root
pos_decisions = clf.decision_function(clf.X_test[clf.y_test==POSTIVE_LABEL])
neg_decisions = clf.decision_function(clf.X_test[clf.y_test==NEGATIVE_LABEL])
means = [numpy.mean(pos_decisions)], [numpy.mean(neg_decisions)]
maxprec = 100.0
precs = [[numpy.min((1/numpy.var(pos_decisions), maxprec))]], [[numpy.min((1/numpy.var(neg_decisions), maxprec))]]
#else:
except (ValueError, AssertionError):
means = [-1.], [1.]
precs = [[1.]], [[1.]]
print("Gaussian params:", term, self.ontology[term]['name'], means, precs)
observed = Mixture(hidden, Gaussian, means, precs)
#observed = ConditionalProbabilityTable([
# ['0', '0', conf[0][0] / posTest], # if term != root else 1.],
# ['0', '1', conf[0][1] / posTest], # if term != root else 0.],
# ['1', '0', conf[1][0] / negTest], # if term != root else 0.],
# ['1', '1', conf[1][1] / negTest]], #if term != root else 1.]],
# [hidden.distribution])
#print("Observed node %s - %s:" % (term, self.ontology[term]['name']))
#print(repr(observed))
#print([p for p in observed.parents if isinstance(p, Stochastic)])
self.ontology[term]['node'][self.fold][self.clfName] = hidden
#self.ontology[term]['clf'][self.fold][self.clfName] = clf, X_validation, y_validation, g_validation
assert self.lenValidation is None or self.lenValidation == len(clf.y_validation)
self.lenValidation = len(clf.y_validation)
self.allobserved[term] = observed
self.allhidden[term] = hidden
self.extranodes.update((p for p in hidden.parents if isinstance(p, Stochastic)))