BayesFlow Stochastic Tensors (contrib)
[TOC]
Classes and helper functions for creating Stochastic Tensors.
StochasticTensor
objects wrap Distribution
objects. Their
values may be samples from the underlying distribution, or the distribution
mean (as governed by value_type
). These objects provide a loss
method for use when sampling from a non-reparameterized distribution.
The loss
method is used in conjunction with stochastic_graph.surrogate_loss
to produce a single differentiable loss in stochastic graphs having
both continuous and discrete stochastic nodes.
Stochastic Tensor Classes
class tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor
Base Class for Tensor-like objects that emit stochastic values.
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.__init__()
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.loss(sample_loss)
Returns the term to add to the surrogate loss.
This method is called by surrogate_loss
. The input sample_loss
should
have already had stop_gradient
applied to it. This is because the
surrogate_loss usually provides a Monte Carlo sample term of the form
differentiable_surrogate * sample_loss
where sample_loss
is considered
constant with respect to the input for purposes of the gradient.
Args:
sample_loss
:Tensor
, sample loss downstream of thisStochasticTensor
.
Returns:
Either None
or a Tensor
.
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.name
tf.contrib.bayesflow.stochastic_tensor.BaseStochasticTensor.value(name=None)
class tf.contrib.bayesflow.stochastic_tensor.StochasticTensor
StochasticTensor is a BaseStochasticTensor backed by a distribution.
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.__init__(dist_cls, name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
Construct a StochasticTensor
.
StochasticTensor
will instantiate a distribution from dist_cls
and
dist_args
and its value
method will return the same value each time
it is called. What value
is returned is controlled by the
dist_value_type
(defaults to SampleAndReshapeValue
).
Some distributions' sample functions are not differentiable (e.g. a sample
from a discrete distribution like a Bernoulli) and so to differentiate
wrt parameters upstream of the sample requires a gradient estimator like
the score function estimator. This is accomplished by passing a
differentiable loss_fn
to the StochasticTensor
, which
defaults to a function whose derivative is the score function estimator.
Calling stochastic_graph.surrogate_loss(final_losses)
will call
loss()
on every StochasticTensor
upstream of final losses.
loss()
will return None for StochasticTensor
s backed by
reparameterized distributions; it will also return None if the value type is
MeanValueType
or if loss_fn=None
.
Args:
dist_cls
: aDistribution
class.name
: a name for thisStochasticTensor
and its ops.dist_value_type
: a_StochasticValueType
, which will determine what thevalue
of thisStochasticTensor
will be. If not provided, the value type set with thevalue_type
context manager will be used.loss_fn
: callable that takes(dt, dt.value(), influenced_loss)
, wheredt
is thisStochasticTensor
, and returns aTensor
loss. By default,loss_fn
is thescore_function
, or more precisely, the integral of the score function, such that when the gradient is taken, the score function results. See thestochastic_gradient_estimators
module for additional loss functions and baselines.**dist_args
: keyword arguments to be passed through todist_cls
on construction.
Raises:
TypeError
: ifdist_cls
is not aDistribution
.TypeError
: ifloss_fn
is notcallable
.
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.graph
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.name
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.StochasticTensor.value_type
Stochastic Tensor Value Types
class tf.contrib.bayesflow.stochastic_tensor.MeanValue
tf.contrib.bayesflow.stochastic_tensor.MeanValue.__init__(stop_gradient=False)
tf.contrib.bayesflow.stochastic_tensor.MeanValue.declare_inputs(unused_stochastic_tensor, unused_inputs_dict)
tf.contrib.bayesflow.stochastic_tensor.MeanValue.popped_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.MeanValue.pushed_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.MeanValue.stop_gradient
class tf.contrib.bayesflow.stochastic_tensor.SampleValue
Draw n samples along a new outer dimension.
This ValueType draws n
samples from StochasticTensors run within its
context, increasing the rank by one along a new outer dimension.
Example:
mu = tf.zeros((2,3))
sigma = tf.ones((2, 3))
with sg.value_type(sg.SampleValue(n=4)):
dt = sg.DistributionTensor(
distributions.Normal, mu=mu, sigma=sigma)
# draws 4 samples each with shape (2, 3) and concatenates
assertEqual(dt.value().get_shape(), (4, 2, 3))
tf.contrib.bayesflow.stochastic_tensor.SampleValue.__init__(n=1, stop_gradient=False)
Sample n
times and concatenate along a new outer dimension.
Args:
n
: A python integer or int32 tensor. The number of samples to take.stop_gradient
: IfTrue
, StochasticTensors' values are wrapped instop_gradient
, to avoid backpropagation through.
tf.contrib.bayesflow.stochastic_tensor.SampleValue.declare_inputs(unused_stochastic_tensor, unused_inputs_dict)
tf.contrib.bayesflow.stochastic_tensor.SampleValue.n
tf.contrib.bayesflow.stochastic_tensor.SampleValue.popped_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.SampleValue.pushed_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.SampleValue.stop_gradient
class tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue
Ask the StochasticTensor for n samples and reshape the result.
Sampling from a StochasticTensor increases the rank of the value by 1 (because each sample represents a new outer dimension).
This ValueType requests n
samples from StochasticTensors run within its
context that the outer two dimensions are reshaped to intermix the samples
with the outermost (usually batch) dimension.
Example:
# mu and sigma are both shaped (2, 3)
mu = [[0.0, -1.0, 1.0], [0.0, -1.0, 1.0]]
sigma = tf.constant([[1.1, 1.2, 1.3], [1.1, 1.2, 1.3]])
with sg.value_type(sg.SampleAndReshapeValue(n=2)):
dt = sg.DistributionTensor(
distributions.Normal, mu=mu, sigma=sigma)
# sample(2) creates a (2, 2, 3) tensor, and the two outermost dimensions
# are reshaped into one: the final value is a (4, 3) tensor.
dt_value = dt.value()
assertEqual(dt_value.get_shape(), (4, 3))
dt_value_val = sess.run([dt_value])[0] # or e.g. run([tf.identity(dt)])[0]
assertEqual(dt_value_val.shape, (4, 3))
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.__init__(n=1, stop_gradient=False)
Sample n
times and reshape the outer 2 axes so rank does not change.
Args:
n
: A python integer or int32 tensor. The number of samples to take.stop_gradient
: IfTrue
, StochasticTensors' values are wrapped instop_gradient
, to avoid backpropagation through.
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.declare_inputs(unused_stochastic_tensor, unused_inputs_dict)
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.n
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.popped_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.pushed_above(unused_value_type)
tf.contrib.bayesflow.stochastic_tensor.SampleAndReshapeValue.stop_gradient
tf.contrib.bayesflow.stochastic_tensor.value_type(dist_value_type)
Creates a value type context for any StochasticTensor created within.
Typical usage:
with sg.value_type(sg.MeanValue(stop_gradients=True)):
dt = sg.DistributionTensor(distributions.Normal, mu=mu, sigma=sigma)
In the example above, dt.value()
(or equivalently, tf.identity(dt)
) will
be the mean value of the Normal distribution, i.e., mu
(possibly
broadcasted to the shape of sigma
). Furthermore, because the MeanValue
was marked with stop_gradients=True
, this value will have been wrapped
in a stop_gradients
call to disable any possible backpropagation.
Args:
dist_value_type
: An instance ofMeanValue
,SampleAndReshapeValue
, or any other stochastic value type.
Yields:
A context for StochasticTensor
objects that controls the
value created when they are initialized.
Raises:
TypeError
: ifdist_value_type
is not an instance of a stochastic value type.
tf.contrib.bayesflow.stochastic_tensor.get_current_value_type()
Automatically Generated StochasticTensors
class tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor
BernoulliTensor
is a StochasticTensor
backed by the distribution Bernoulli
.
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.name
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.BernoulliTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor
BernoulliWithSigmoidPTensor
is a StochasticTensor
backed by the distribution BernoulliWithSigmoidP
.
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.name
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.BernoulliWithSigmoidPTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.BetaTensor
BetaTensor
is a StochasticTensor
backed by the distribution Beta
.
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.name
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.BetaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor
BetaWithSoftplusABTensor
is a StochasticTensor
backed by the distribution BetaWithSoftplusAB
.
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.name
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.BetaWithSoftplusABTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.BinomialTensor
BinomialTensor
is a StochasticTensor
backed by the distribution Binomial
.
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.graph
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.name
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.BinomialTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor
CategoricalTensor
is a StochasticTensor
backed by the distribution Categorical
.
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.graph
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.name
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.CategoricalTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor
Chi2Tensor
is a StochasticTensor
backed by the distribution Chi2
.
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.distribution
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.dtype
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.graph
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.name
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.Chi2Tensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor
Chi2WithAbsDfTensor
is a StochasticTensor
backed by the distribution Chi2WithAbsDf
.
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.graph
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.name
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.Chi2WithAbsDfTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.DirichletTensor
DirichletTensor
is a StochasticTensor
backed by the distribution Dirichlet
.
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.graph
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.name
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.DirichletTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor
DirichletMultinomialTensor
is a StochasticTensor
backed by the distribution DirichletMultinomial
.
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.graph
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.name
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.DirichletMultinomialTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor
ExponentialTensor
is a StochasticTensor
backed by the distribution Exponential
.
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.graph
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.name
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.ExponentialTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor
ExponentialWithSoftplusLamTensor
is a StochasticTensor
backed by the distribution ExponentialWithSoftplusLam
.
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.graph
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.name
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.ExponentialWithSoftplusLamTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.GammaTensor
GammaTensor
is a StochasticTensor
backed by the distribution Gamma
.
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.name
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.GammaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor
GammaWithSoftplusAlphaBetaTensor
is a StochasticTensor
backed by the distribution GammaWithSoftplusAlphaBeta
.
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.name
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.GammaWithSoftplusAlphaBetaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor
InverseGammaTensor
is a StochasticTensor
backed by the distribution InverseGamma
.
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.name
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor
InverseGammaWithSoftplusAlphaBetaTensor
is a StochasticTensor
backed by the distribution InverseGammaWithSoftplusAlphaBeta
.
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.name
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.InverseGammaWithSoftplusAlphaBetaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor
LaplaceTensor
is a StochasticTensor
backed by the distribution Laplace
.
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.graph
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.name
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.LaplaceTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor
LaplaceWithSoftplusScaleTensor
is a StochasticTensor
backed by the distribution LaplaceWithSoftplusScale
.
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.graph
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.name
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.LaplaceWithSoftplusScaleTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MixtureTensor
MixtureTensor
is a StochasticTensor
backed by the distribution Mixture
.
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.name
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MixtureTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor
MultinomialTensor
is a StochasticTensor
backed by the distribution Multinomial
.
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultinomialTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor
MultivariateNormalCholeskyTensor
is a StochasticTensor
backed by the distribution MultivariateNormalCholesky
.
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalCholeskyTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor
MultivariateNormalDiagTensor
is a StochasticTensor
backed by the distribution MultivariateNormalDiag
.
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor
MultivariateNormalDiagPlusVDVTTensor
is a StochasticTensor
backed by the distribution MultivariateNormalDiagPlusVDVT
.
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagPlusVDVTTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor
MultivariateNormalDiagWithSoftplusStDevTensor
is a StochasticTensor
backed by the distribution MultivariateNormalDiagWithSoftplusStDev
.
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalDiagWithSoftplusStDevTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor
MultivariateNormalFullTensor
is a StochasticTensor
backed by the distribution MultivariateNormalFull
.
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.graph
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.name
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.MultivariateNormalFullTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.NormalTensor
NormalTensor
is a StochasticTensor
backed by the distribution Normal
.
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.graph
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.name
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.NormalTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor
NormalWithSoftplusSigmaTensor
is a StochasticTensor
backed by the distribution NormalWithSoftplusSigma
.
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.name
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.NormalWithSoftplusSigmaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.PoissonTensor
PoissonTensor
is a StochasticTensor
backed by the distribution Poisson
.
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.graph
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.name
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.PoissonTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor
QuantizedDistributionTensor
is a StochasticTensor
backed by the distribution QuantizedDistribution
.
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.graph
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.name
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.QuantizedDistributionTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.StudentTTensor
StudentTTensor
is a StochasticTensor
backed by the distribution StudentT
.
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.graph
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.name
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.StudentTTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor
StudentTWithAbsDfSoftplusSigmaTensor
is a StochasticTensor
backed by the distribution StudentTWithAbsDfSoftplusSigma
.
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.graph
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.name
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.StudentTWithAbsDfSoftplusSigmaTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor
TransformedDistributionTensor
is a StochasticTensor
backed by the distribution TransformedDistribution
.
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.graph
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.name
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.TransformedDistributionTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.UniformTensor
UniformTensor
is a StochasticTensor
backed by the distribution Uniform
.
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.graph
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.name
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.UniformTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor
WishartCholeskyTensor
is a StochasticTensor
backed by the distribution WishartCholesky
.
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.graph
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.name
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.WishartCholeskyTensor.value_type
class tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor
WishartFullTensor
is a StochasticTensor
backed by the distribution WishartFull
.
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.__init__(name=None, dist_value_type=None, loss_fn=score_function, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.clone(name=None, **dist_args)
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.distribution
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.dtype
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.entropy(name='entropy')
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.graph
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.input_dict
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.loss(final_loss, name='Loss')
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.mean(name='mean')
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.name
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.value(name='value')
tf.contrib.bayesflow.stochastic_tensor.WishartFullTensor.value_type
Other Functions and Classes
class tf.contrib.bayesflow.stochastic_tensor.ObservedStochasticTensor
A StochasticTensor with an observed value.
tf.contrib.bayesflow.stochastic_tensor.ObservedStochasticTensor.__init__(dist_cls, value, name=None, **dist_args)
Construct an ObservedStochasticTensor
.
ObservedStochasticTensor
will instantiate a distribution from dist_cls
and dist_args
but use the provided value instead of sampling from the
distribution. The provided value argument must be appropriately shaped
to have come from the constructed distribution.
Args:
dist_cls
: aDistribution
class.value
: a Tensor containing the observed valuename
: a name for thisObservedStochasticTensor
and its ops.**dist_args
: keyword arguments to be passed through todist_cls
on construction.
Raises:
TypeError
: ifdist_cls
is not aDistribution
.ValueError
: ifvalue
is not compatible with the distribution.