tfp.experimental.distributions.IncrementLogProb

A distribution-like object representing an unnormalized density at a point.

IncrementLogProb is a distribution-like class that represents an unnormalized density, sometimes called a "factor". Its raison d'être is to provide an overall offset to the log probability of a JointDensity.

It has no log_prob method, instead surfacing the unnormalized log probability/density through its unnormalized_log_prob method. It retains a sample method, which returns an empty tensor with the same dtype as the increment_log_prob argument provided to it originally.

log_prob_increment Log probability/density to increment by.
validate_args This argument is ignored, but is present because it is used in certain situations where Distributions are expected.
allow_nan_stats This argument is ignored, but is present because it is used in certain situations where Distributions are expected.
reparameterization_type This argument is ignored, but is present because it is used in certain situations where Distributions are expected.
name Python str name prefixed to Ops created by this class.

allow_nan_stats

batch_shape

dtype

event_shape

experimental_shard_axis_names The list or structure of lists of active shard axis names.
log_prob_increment The amount to increment log probability by.
name

parameters

reparameterization_type

Methods

batch_shape_tensor

View source

Shape of a single sample from a single event index as a 1-D Tensor.

The batch dimensions are indexes into independent, non-identical parameterizations of this distribution.

Args
name name to give to the op

Returns
batch_shape Tensor.

event_shape_tensor

View source

Shape of a single sample from a single batch as a 1-D int32 Tensor.

Args
name name to give to the op

Returns
event_shape Tensor.

sample

View source

unnormalized_log_prob

View source