def test_tt(self):
true_sampler = theano.function([self.coin_weight], self.coin)
sample, ll, updates = mh_sample(self.s_rng, [self.coin_weight])
sampler = theano.function([self.coin], sample, updates=updates)
for i in range(100):
print sampler(true_sampler(0.9))
def test_tt(self):
true_sampler = theano.function([self.coin_weight], self.coin)
sample, ll, updates = mh_sample(self.s_rng, [self.coin_weight])
sampler = theano.function([self.coin], sample, updates=updates)
for i in range(100):
print sampler(true_sampler(0.9))
def test_mh_sample(self):
sample, ll, updates = mh_sample(self.s_rng, [self.M, self.V], observations={self.X: self.X_data}, lag = 100)
sampler = theano.function([], sample, updates=updates)
data = []
for i in range(100):
print i
data.append(sampler())
pylab.subplot(211)
pylab.hist(numpy.asarray(data)[:,0])
pylab.subplot(212)
pylab.hist(numpy.asarray(data)[:,1])
pylab.show()
def test_mh_sample(self):
sample, ll, updates = mh_sample(self.s_rng, [self.M, self.V],
observations={self.X: self.X_data},
lag=100)
sampler = theano.function([], sample, updates=updates)
data = []
for i in range(100):
print i
data.append(sampler())
pylab.subplot(211)
pylab.hist(numpy.asarray(data)[:, 0])
pylab.subplot(212)
pylab.hist(numpy.asarray(data)[:, 1])
pylab.show()
[0,0,1,0],
[0,0,1,0],
[0,0,1,0]], dtype=theano.config.floatX)
document_3 = numpy.asarray([[1,0,0,0],
[0,0,0,1],
[0,1,0,0],
[0,1,0,0],
[0,0,1,0],
[0,1,0,0],
[0,0,0,1],
[1,0,0,0],
[0,0,1,0],
[0,0,1,0]], dtype=theano.config.floatX)
# Map documents to RVs
givens = {get_words(1, 10): document_1,
get_words(2, 10): document_2,
get_words(3, 10): document_3}
# Build sampler
sample, ll, updates = mh_sample(s_rng, [doc_mixture(1), doc_mixture(2), doc_mixture(3), topic_mixture(0), topic_mixture(1)])
sampler = theano.function([], sample, updates=updates, givens=givens, allow_input_downcast=True)
# Run sampling
for i in range(10000):
d = sampler()
if i % 1000 == 0:
print d
[[1, 0, 0, 0], [0, 0, 0, 1], [0, 1, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0],
[0, 1, 0, 0], [0, 0, 0, 1], [1, 0, 0, 0], [0, 0, 1, 0], [0, 0, 1, 0]],
dtype=theano.config.floatX)
# Map documents to RVs
givens = {
get_words(1, 10): document_1,
get_words(2, 10): document_2,
get_words(3, 10): document_3
}
# Build sampler
sample, ll, updates = mh_sample(s_rng, [
doc_mixture(1),
doc_mixture(2),
doc_mixture(3),
topic_mixture(0),
topic_mixture(1)
])
sampler = theano.function([],
sample,
updates=updates,
givens=givens,
allow_input_downcast=True)
# Run sampling
for i in range(10000):
d = sampler()
if i % 1000 == 0:
print d