tf.contrib.distributions.VectorLaplaceDiag

The vectorization of the Laplace distribution on R^k.

tf.contrib.distributions.VectorLaplaceDiag(
    loc=None, scale_diag=None, scale_identity_multiplier=None, validate_args=False,
    allow_nan_stats=True, name='VectorLaplaceDiag'
)

The vector laplace distribution is defined over R^k, and parameterized by a (batch of) length-k loc vector (the means) and a (batch of) k x k scale matrix: covariance = 2 * scale @ scale.T, where @ denotes matrix-multiplication.

Mathematical Details

The probability density function (pdf) is,

pdf(x; loc, scale) = exp(-||y||_1) / Z,
y = inv(scale) @ (x - loc),
Z = 2**k |det(scale)|,

where:

loc is a vector in R^k,
scale is a linear operator in R^{k x k}, cov = scale @ scale.T,
Z denotes the normalization constant, and,
||y||_1 denotes the l1 norm of y, `sum_i |y_i|.

A (non-batch) scale matrix is:

scale = diag(scale_diag + scale_identity_multiplier * ones(k))

where:

scale_diag.shape = [k], and,
scale_identity_multiplier.shape = [].

Additional leading dimensions (if any) will index batches.

If both scale_diag and scale_identity_multiplier are None, then scale is the Identity matrix.

The VectorLaplace distribution is a member of the location-scale family, i.e., it can be constructed as,

X = (X_1, ..., X_k), each X_i ~ Laplace(loc=0, scale=1)
Y = (Y_1, ...,Y_k) = scale @ X + loc

About `VectorLaplace` and `Vector` distributions in TensorFlow.

The VectorLaplace is a non-standard distribution that has useful properties.

The marginals Y_1, ..., Y_k are not Laplace random variables, due to the fact that the sum of Laplace random variables is not Laplace.

Instead, Y is a vector whose components are linear combinations of Laplace random variables. Thus, Y lives in the vector space generated by vectors of Laplace distributions. This allows the user to decide the mean and covariance (by setting loc and scale), while preserving some properties of the Laplace distribution. In particular, the tails of Y_i will be (up to polynomial factors) exponentially decaying.

To see this last statement, note that the pdf of Y_i is the convolution of the pdf of k independent Laplace random variables. One can then show by induction that distributions with exponential (up to polynomial factors) tails are closed under convolution.

Examples

import tensorflow_probability as tfp
tfd = tfp.distributions

# Initialize a single 2-variate VectorLaplace.
vla = tfd.VectorLaplaceDiag(
    loc=[1., -1],
    scale_diag=[1, 2.])

vla.mean().eval()
# ==> [1., -1]

vla.stddev().eval()
# ==> [1., 2] * sqrt(2)

# Evaluate this on an observation in `R^2`, returning a scalar.
vla.prob([-1., 0]).eval()  # shape: []

# Initialize a 3-batch, 2-variate scaled-identity VectorLaplace.
vla = tfd.VectorLaplaceDiag(
    loc=[1., -1],
    scale_identity_multiplier=[1, 2., 3])

vla.mean().eval()  # shape: [3, 2]
# ==> [[1., -1]
#      [1, -1],
#      [1, -1]]

vla.stddev().eval()  # shape: [3, 2]
# ==> sqrt(2) * [[1., 1],
#                [2, 2],
#                [3, 3]]

# Evaluate this on an observation in `R^2`, returning a length-3 vector.
vla.prob([-1., 0]).eval()  # shape: [3]

# Initialize a 2-batch of 3-variate VectorLaplace's.
vla = tfd.VectorLaplaceDiag(
    loc=[[1., 2, 3],
         [11, 22, 33]]           # shape: [2, 3]
    scale_diag=[[1., 2, 3],
                [0.5, 1, 1.5]])  # shape: [2, 3]

# Evaluate this on a two observations, each in `R^3`, returning a length-2
# vector.
x = [[-1., 0, 1],
     [-11, 0, 11.]]   # shape: [2, 3].
vla.prob(x).eval()    # shape: [2]

Args
`loc`	Floating-point `Tensor`. If this is set to `None`, `loc` is implicitly `0`. When specified, may have shape `[B1, ..., Bb, k]` where `b >= 0` and `k` is the event size.
`scale_diag`	Non-zero, floating-point `Tensor` representing a diagonal matrix added to `scale`. May have shape `[B1, ..., Bb, k]`, `b >= 0`, and characterizes `b`-batches of `k x k` diagonal matrices added to `scale`. When both `scale_identity_multiplier` and `scale_diag` are `None` then `scale` is the `Identity`.
`scale_identity_multiplier`	Non-zero, floating-point `Tensor` representing a scaled-identity-matrix added to `scale`. May have shape `[B1, ..., Bb]`, `b >= 0`, and characterizes `b`-batches of scaled `k x k` identity matrices added to `scale`. When both `scale_identity_multiplier` and `scale_diag` are `None` then `scale` is the `Identity`.
`validate_args`	Python `bool`, default `False`. When `True` distribution parameters are checked for validity despite possibly degrading runtime performance. When `False` invalid inputs may silently render incorrect outputs.
`allow_nan_stats`	Python `bool`, default `True`. When `True`, statistics (e.g., mean, mode, variance) use the value "`NaN`" to indicate the result is undefined. When `False`, an exception is raised if one or more of the statistic's batch members are undefined.
`name`	Python `str` name prefixed to Ops created by this class.

Raises
`ValueError`	if at most `scale_identity_multiplier` is specified.

Attributes
`allow_nan_stats`	Python `bool` describing behavior when a stat is undefined. Stats return +/- infinity when it makes sense. E.g., the variance of a Cauchy distribution is infinity. However, sometimes the statistic is undefined, e.g., if a distribution's pdf does not achieve a maximum within the support of the distribution, the mode is undefined. If the mean is undefined, then by definition the variance is undefined. E.g. the mean for Student's T for df = 1 is undefined (no clear way to say it is either + or - infinity), so the variance = E[(X - mean)**2] is also undefined.
`batch_shape`	Shape of a single sample from a single event index as a `TensorShape`. May be partially defined or unknown. The batch dimensions are indexes into independent, non-identical parameterizations of this distribution.
`bijector`	Function transforming x => y.
`distribution`	Base distribution, p(x).
`dtype`	The `DType` of `Tensor`s handled by this `Distribution`.
`event_shape`	Shape of a single sample from a single batch as a `TensorShape`. May be partially defined or unknown.
`loc`	The `loc` `Tensor` in `Y = scale @ X + loc`.
`name`	Name prepended to all ops created by this `Distribution`.
`parameters`	Dictionary of parameters used to instantiate this `Distribution`.
`reparameterization_type`	Describes how samples from the distribution are reparameterized. Currently this is one of the static instances `distributions.FULLY_REPARAMETERIZED` or `distributions.NOT_REPARAMETERIZED`.
`scale`	The `scale` `LinearOperator` in `Y = scale @ X + loc`.
`validate_args`	Python `bool` indicating possibly expensive checks are enabled.

Args
`value`	`float` or `double` `Tensor`.
`name`	Python `str` prepended to names of ops created by this function.

Args
`other`	`tfp.distributions.Distribution` instance.
`name`	Python `str` prepended to names of ops created by this function.

Args
`sample_shape`	`Tensor` or python list/tuple. Desired shape of a call to `sample()`.
`name`	name to prepend ops with.

Args
`sample_shape`	0D or 1D `int32` `Tensor`. Shape of the generated samples.
`seed`	Python integer seed for RNG
`name`	name to give to the op.

tf.contrib.distributions.VectorLaplaceDiag

Mathematical Details

About VectorLaplace and Vector distributions in TensorFlow.

Examples

Methods

batch_shape_tensor

cdf

copy

covariance

cross_entropy

entropy

event_shape_tensor

is_scalar_batch

is_scalar_event

kl_divergence

log_cdf

log_prob

log_survival_function

mean

mode

param_shapes

param_static_shapes

prob

quantile

sample

stddev

survival_function

variance

About `VectorLaplace` and `Vector` distributions in TensorFlow.

`batch_shape_tensor`

`cdf`

`copy`

`covariance`

`cross_entropy`

`entropy`

`event_shape_tensor`

`is_scalar_batch`

`is_scalar_event`

`kl_divergence`

`log_cdf`

`log_prob`

`log_survival_function`

`mean`

`mode`

`param_shapes`

`param_static_shapes`

`prob`

`quantile`

`sample`

`stddev`

`survival_function`

`variance`