tensorplay.distributions API

Functions 2

kl_divergence

tensorplay.distributions.kl_divergence(p: Distribution, q: Distribution) → Tensor[source]

Compute Kullback-Leibler divergence $KL(p \| q)$ between two distributions.

KL(p \| q) = \int p(x) \log\frac {p(x)} {q(x)} \,dx

Parameters:

p (Distribution) – A Distribution object.
q (Distribution) – A Distribution object.

Returns:

A batch of KL divergences of shape batch_shape.

Return type:

Tensor

Raises:

NotImplementedError – If the distribution types have not been registered via register_kl().

KL divergence is currently implemented for the following distribution pairs:

Bernoulli and Bernoulli
Bernoulli and Poisson
Beta and Beta
Beta and ContinuousBernoulli
Beta and Exponential
Beta and Gamma
Beta and Normal
Beta and Pareto
Beta and Uniform
Binomial and Binomial
Categorical and Categorical
Cauchy and Cauchy
ContinuousBernoulli and ContinuousBernoulli
ContinuousBernoulli and Exponential
ContinuousBernoulli and Normal
ContinuousBernoulli and Pareto
ContinuousBernoulli and Uniform
Dirichlet and Dirichlet
Exponential and Beta
Exponential and ContinuousBernoulli
Exponential and Exponential
Exponential and Gamma
Exponential and Gumbel
Exponential and Normal
Exponential and Pareto
Exponential and Uniform
ExponentialFamily and ExponentialFamily
Gamma and Beta
Gamma and ContinuousBernoulli
Gamma and Exponential
Gamma and Gamma
Gamma and Gumbel
Gamma and Normal
Gamma and Pareto
Gamma and Uniform
Geometric and Geometric
Gumbel and Beta
Gumbel and ContinuousBernoulli
Gumbel and Exponential
Gumbel and Gamma
Gumbel and Gumbel
Gumbel and Normal
Gumbel and Pareto
Gumbel and Uniform
HalfNormal and HalfNormal
Independent and Independent
Laplace and Beta
Laplace and ContinuousBernoulli
Laplace and Exponential
Laplace and Gamma
Laplace and Laplace
Laplace and Normal
Laplace and Pareto
Laplace and Uniform
LowRankMultivariateNormal and LowRankMultivariateNormal
LowRankMultivariateNormal and MultivariateNormal
MultivariateNormal and LowRankMultivariateNormal
MultivariateNormal and MultivariateNormal
Normal and Beta
Normal and ContinuousBernoulli
Normal and Exponential
Normal and Gamma
Normal and Gumbel
Normal and Laplace
Normal and Normal
Normal and Pareto
Normal and Uniform
OneHotCategorical and OneHotCategorical
Pareto and Beta
Pareto and ContinuousBernoulli
Pareto and Exponential
Pareto and Gamma
Pareto and Normal
Pareto and Pareto
Pareto and Uniform
Poisson and Bernoulli
Poisson and Binomial
Poisson and Poisson
TransformedDistribution and TransformedDistribution
Uniform and Beta
Uniform and ContinuousBernoulli
Uniform and Exponential
Uniform and Gamma
Uniform and Gumbel
Uniform and Normal
Uniform and Pareto
Uniform and Uniform

register_kl

functionFull reference ↗

tensorplay.distributions.register_kl(type_p, type_q)[source]

Decorator to register a pairwise function with kl_divergence(). Usage:

@register_kl(Normal, Normal)
def kl_normal_normal(p, q):
    # insert implementation here

Lookup returns the most specific (type,type) match ordered by subclass. If the match is ambiguous, a RuntimeWarning is raised. For example to resolve the ambiguous situation:

@register_kl(BaseP, DerivedQ)
def kl_version1(p, q): ...
@register_kl(DerivedP, BaseQ)
def kl_version2(p, q): ...

you should register a third most-specific implementation, e.g.:

register_kl(DerivedP, DerivedQ)(kl_version1)  # Break the tie.

Parameters:

type_p (type) – A subclass of Distribution.
type_q (type) – A subclass of Distribution.

Classes 62

AbsTransform

classFull reference ↗

class tensorplay.distributions.AbsTransform(cache_size: int = 0)[source]

Transform via the mapping $y = |x|$ .

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

log_abs_det_jacobian(x, y): Computes the log det jacobian log |dy/dx| given input and output.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

AffineTransform

classFull reference ↗

class tensorplay.distributions.AffineTransform(loc: Tensor | float, scale: Tensor | float, event_dim: int = 0, cache_size: int = 0)[source]

Transform via the pointwise affine mapping $y = \text{loc} + \text{scale} \times x$ .

Parameters:

loc (Tensor or float) – Location parameter.
scale (Tensor or float) – Scale parameter.
event_dim (int) – Optional size of event_shape. This should be zero for univariate random variables, 1 for distributions over vectors, 2 for distributions over matrices, etc.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

Bernoulli

classFull reference ↗

Creates a Bernoulli distribution parameterized by probs or logits (but not both).

Samples are binary (0 or 1). They take the value 1 with probability p and 0 with probability 1 - p.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Bernoulli(tensorplay.tensor([0.3]))
>>> m.sample()  # 30% chance 1; 70% chance 0
tensor([ 0.])

Parameters:

probs (Number, Tensor) – the probability of sampling 1
logits (Number, Tensor) – the log-odds of sampling 1
validate_args (bool, optional) – whether to validate arguments, None by default

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

The default behavior mimics Python’s assert statement: validation is on by default, but is disabled if Python is run in optimized mode (via python -O). Validation may be expensive, so you may want to disable it once a model is working.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Beta

classFull reference ↗

class tensorplay.distributions.Beta(concentration1: Tensor | float, concentration0: Tensor | float, validate_args: bool | None = None)[source]

Beta distribution parameterized by concentration1 and concentration0.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Beta(tensorplay.tensor([0.5]), tensorplay.tensor([0.5]))
>>> m.sample()  # Beta distributed with concentration concentration1 and concentration0
tensor([ 0.1046])

Parameters:

concentration1 (float or Tensor) – 1st concentration parameter of the distribution (often referred to as alpha)
concentration0 (float or Tensor) – 2nd concentration parameter of the distribution (often referred to as beta)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

Returns tensor containing all values supported by a discrete distribution. The result will enumerate over dimension 0, so the shape of the result will be (cardinality,) + batch_shape + event_shape (where event_shape = () for univariate distributions).

Note that this enumerates over all batched tensors in lock-step [[0, 0], [1, 1], …]. With expand=False, enumeration happens along dim 0, but with the remaining batch dimensions being singleton dimensions, [[0], [1], ...

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Binomial

classFull reference ↗

class tensorplay.distributions.Binomial(total_count: Tensor | int = 1, probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a Binomial distribution parameterized by total_count and either probs or logits (but not both). total_count must be broadcastable with probs/logits.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Binomial(100, tensorplay.tensor([0 , .2, .8, 1]))
>>> x = m.sample()
tensor([   0.,   22.,   71.,  100.])

>>> m = Binomial(tensorplay.tensor([[5.], [10.]]), tensorplay.tensor([0.5, 0.8]))
>>> x = m.sample()
tensor([[ 4.,  5.],
        [ 7.,  6.]])

Parameters:

total_count (int or Tensor) – number of Bernoulli trials
probs (Tensor) – Event probabilities
logits (Tensor) – Event log-odds

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Categorical

classFull reference ↗

class tensorplay.distributions.Categorical(probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a categorical distribution parameterized by either probs or logits (but not both).

Note

It is equivalent to the distribution that tensorplay.multinomial() samples from.

Samples are integers from $\{0, \ldots, K-1\}$ where K is probs.size(-1).

If probs is 1-dimensional with length-K, each element is the relative probability of sampling the class at that index.

If probs is N-dimensional, the first N-1 dimensions are treated as a batch of relative probability vectors.

Note

The probs argument must be non-negative, finite and have a non-zero sum, and it will be normalized to sum to 1 along the last dimension. probs will return this normalized value. The logits argument will be interpreted as unnormalized log probabilities and can therefore be any real number. It will likewise be normalized so that the resulting probabilities sum to 1 along the last dimension. logits will return this normalized value.

CatTransform

classFull reference ↗

class tensorplay.distributions.CatTransform(tseq: Sequence[Transform], dim: int = 0, lengths: Sequence[int] | None = None, cache_size: int = 0)[source]

Transform functor that applies a sequence of transforms tseq component-wise to each submatrix at dim, of length lengths[dim], in a way compatible with tensorplay.cat().

Example:

x0 = tensorplay.cat([tensorplay.range(1, 10), tensorplay.range(1, 10)], dim=0)
x = tensorplay.cat([x0, x0], dim=0)
t0 = CatTransform([ExpTransform(), identity_transform], dim=0, lengths=[10, 10])
t = CatTransform([t0, t0], dim=0, lengths=[20, 20])
y = t(x)

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

Cauchy

classFull reference ↗

class tensorplay.distributions.Cauchy(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Samples from a Cauchy (Lorentz) distribution. The distribution of the ratio of independent normally distributed random variables with means 0 follows a Cauchy distribution.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Cauchy(tensorplay.tensor([0.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # sample from a Cauchy distribution with loc=0 and scale=1
tensor([ 2.3214])

Parameters:

loc (float or Tensor) – mode or median of the distribution.
scale (float or Tensor) – half width at half maximum.

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Chi2

classFull reference ↗

class tensorplay.distributions.Chi2(df: Tensor | float, validate_args: bool | None = None)[source]

Creates a Chi-squared distribution parameterized by shape parameter df. This is exactly equivalent to Gamma(alpha=0.5*df, beta=0.5)

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Chi2(tensorplay.tensor([1.0]))
>>> m.sample()  # Chi2 distributed with shape df=1
tensor([ 0.1046])

Parameters:: df (float or Tensor) – shape parameter of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

ComposeTransform

classFull reference ↗

class tensorplay.distributions.ComposeTransform(parts: list[Transform], cache_size: int = 0)[source]

Composes multiple transforms in a chain. The transforms being composed are responsible for caching.

Parameters:

parts (list of Transform) – A list of transforms to compose.
cache_size (int) – Size of cache. If zero, no caching is done. If one, the latest single value is cached. Only 0 and 1 are supported.

ContinuousBernoulli

classFull reference ↗

Creates a continuous Bernoulli distribution parameterized by probs or logits (but not both).

The distribution is supported in [0, 1] and parameterized by ‘probs’ (in (0,1)) or ‘logits’ (real-valued). Note that, unlike the Bernoulli, ‘probs’ does not correspond to a probability and ‘logits’ does not correspond to log-odds, but the same names are used due to the similarity with the Bernoulli. See [1] for more details.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = ContinuousBernoulli(tensorplay.tensor([0.3]))
>>> m.sample()
tensor([ 0.2538])

Parameters:

probs (Number, Tensor) – (0,1) valued parameters
logits (Number, Tensor) – real valued parameters whose sigmoid matches ‘probs’

[1] The continuous Bernoulli: fixing a pervasive error in variational autoencoders, Loaiza-Ganem G and Cunningham JP, NeurIPS 2019. https://arxiv.org/abs/1907.06845

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

CorrCholeskyTransform

classFull reference ↗

class tensorplay.distributions.CorrCholeskyTransform(cache_size: int = 0)[source]

Transforms an unconstrained real vector $x$ with length $D*(D-1)/2$ into the Cholesky factor of a D-dimension correlation matrix. This Cholesky factor is a lower triangular matrix with positive diagonals and unit Euclidean norm for each row. The transform is processed as follows:

First we convert x into a lower triangular matrix in row order.

For each row $X_i$ of the lower triangular part, we apply a signed version of class StickBreakingTransform to transform $X_i$ into a unit Euclidean length vector using the following steps: - Scales into the interval $(-1, 1)$ domain: $r_i = \tanh(X_i)$ . - Transforms into an unsigned domain: $z_i = r_i^2$ . - Applies $s_i = StickBreakingTransform(z_i)$ . - Transforms back into signed domain: $y_i = sign(r_i) * \sqrt{s_i}$ .

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

CumulativeDistributionTransform

classFull reference ↗

class tensorplay.distributions.CumulativeDistributionTransform(distribution: Distribution, cache_size: int = 0)[source]

Transform via the cumulative distribution function of a probability distribution.

Parameters:: distribution (Distribution) – Distribution whose cumulative distribution function to use for the transformation.

Example:

# Construct a Gaussian copula from a multivariate normal.
base_dist = MultivariateNormal(
    loc=tensorplay.zeros(2),
    scale_tril=LKJCholesky(2).sample(),
)
transform = CumulativeDistributionTransform(Normal(0, 1))
copula = TransformedDistribution(base_dist, [transform])

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

Dirichlet

classFull reference ↗

class tensorplay.distributions.Dirichlet(concentration: Tensor, validate_args: bool | None = None)[source]

Creates a Dirichlet distribution parameterized by concentration concentration.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Dirichlet(tensorplay.tensor([0.5, 0.5]))
>>> m.sample()  # Dirichlet distributed with concentration [0.5, 0.5]
tensor([ 0.1046,  0.8954])

Parameters:: concentration (Tensor) – concentration parameter of the distribution (often referred to as alpha)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Distribution

classFull reference ↗

class tensorplay.distributions.Distribution(batch_shape: Size = (), event_shape: Size = (), validate_args: bool | None = None)[source]

Distribution is the abstract base class for probability distributions.

Parameters:

batch_shape (tensorplay.Size) – The shape over which parameters are batched.
event_shape (tensorplay.Size) – The shape of a single sample (without batching).
validate_args (bool, optional) – Whether to validate arguments. Default: None.

property arg_constraints: dict[str, Constraint]: Returns a dictionary from argument names to Constraint objects that should be satisfied by each argument of this distribution. Args that are not tensors need not appear in this dict.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor[source]

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy() → Tensor[source]

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor[source]

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

expand(batch_shape: Size | Sequence[int], _instance=None)[source]

Returns a new distribution instance (or populates an existing instance provided by a derived class) with batch dimensions expanded to batch_shape. This method calls expand on the distribution’s parameters. As such, this does not allocate new memory for the expanded distribution instance. Additionally, this does not repeat any args checking or parameter broadcasting in __init__.py, when an instance is first created.

Parameters:

batch_shape (tensorplay.Size) – the desired expanded size.
_instance – new instance provided by subclasses that need to override .expand.

Returns:

New distribution instance with batch dimensions expanded to batch_shape.

icdf(value: Tensor) → Tensor[source]

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

log_prob(value: Tensor) → Tensor[source]

Returns the log of the probability density/mass function evaluated at value.

Parameters:: value (Tensor)

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor[source]

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor[source]: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor[source]: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor[source]: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None[source]

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property support: Constraint | None: Returns a Constraint object representing this distribution’s support.

property variance: Tensor: Returns the variance of the distribution.

Exponential

classFull reference ↗

class tensorplay.distributions.Exponential(rate: Tensor | float, validate_args: bool | None = None)[source]

Creates an Exponential distribution parameterized by rate.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Exponential(tensorplay.tensor([1.0]))
>>> m.sample()  # Exponential distributed with rate=1
tensor([ 0.1046])

Parameters:: rate (float or Tensor) – rate = 1 / scale of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

ExponentialFamily

classFull reference ↗

class tensorplay.distributions.ExponentialFamily(batch_shape: Size = (), event_shape: Size = (), validate_args: bool | None = None)[source]

ExponentialFamily is the abstract base class for probability distributions belonging to an exponential family, whose probability mass/density function is defined below

p_{F}(x; \theta) = \exp(\langle t(x), \theta\rangle - F(\theta) + k(x))

where $\theta$ denotes the natural parameters, $t(x)$ denotes the sufficient statistic, $F(\theta)$ is the log normalizer function for a given family and $k(x)$ is the carrier measure.

Note

This class is an intermediary between the Distribution class and distributions which belong to an exponential family mainly to check the correctness of the .entropy() and analytic KL divergence methods. We use this class to compute the entropy and KL divergence using the AD framework and Bregman divergences (courtesy of: Frank Nielsen and Richard Nock, Entropies and Cross-entropies of Exponential Families).

property arg_constraints: dict[str, Constraint]: Returns a dictionary from argument names to Constraint objects that should be satisfied by each argument of this distribution. Args that are not tensors need not appear in this dict.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy()[source]: Method to compute the entropy using Bregman divergence of the log normalizer.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

expand(batch_shape: Size | Sequence[int], _instance=None)

Parameters:

batch_shape (tensorplay.Size) – the desired expanded size.
_instance – new instance provided by subclasses that need to override .expand.

Returns:

New distribution instance with batch dimensions expanded to batch_shape.

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

log_prob(value: Tensor) → Tensor

Returns the log of the probability density/mass function evaluated at value.

Parameters:: value (Tensor)

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property support: Constraint | None: Returns a Constraint object representing this distribution’s support.

property variance: Tensor: Returns the variance of the distribution.

ExpTransform

classFull reference ↗

class tensorplay.distributions.ExpTransform(cache_size: int = 0)[source]

Transform via the mapping $y = \exp(x)$ .

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

FisherSnedecor

classFull reference ↗

class tensorplay.distributions.FisherSnedecor(df1: Tensor | float, df2: Tensor | float, validate_args: bool | None = None)[source]

Creates a Fisher-Snedecor distribution parameterized by df1 and df2.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = FisherSnedecor(tensorplay.tensor([1.0]), tensorplay.tensor([2.0]))
>>> m.sample()  # Fisher-Snedecor-distributed with df1=1 and df2=2
tensor([ 0.2453])

Parameters:

df1 (float or Tensor) – degrees of freedom parameter 1
df2 (float or Tensor) – degrees of freedom parameter 2

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Gamma

classFull reference ↗

class tensorplay.distributions.Gamma(concentration: Tensor | float, rate: Tensor | float, validate_args: bool | None = None)[source]

Creates a Gamma distribution parameterized by shape concentration and rate.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Gamma(tensorplay.tensor([1.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # Gamma distributed with concentration=1 and rate=1
tensor([ 0.1046])

Parameters:

concentration (float or Tensor) – shape parameter of the distribution (often referred to as alpha)
rate (float or Tensor) – rate parameter of the distribution (often referred to as beta), rate = 1 / scale

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

GeneralizedPareto

classFull reference ↗

class tensorplay.distributions.GeneralizedPareto(loc, scale, concentration, validate_args=None)[source]

Creates a Generalized Pareto distribution parameterized by loc, scale, and concentration.

The Generalized Pareto distribution is a family of continuous probability distributions on the real line. Special cases include Exponential (when loc = 0, concentration = 0), Pareto (when concentration > 0, loc = scale / concentration), and Uniform (when concentration = -1).

This distribution is often used to model the tails of other distributions. This implementation is based on the implementation in TensorFlow Probability.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = GeneralizedPareto(tensorplay.tensor([0.1]), tensorplay.tensor([2.0]), tensorplay.tensor([0.4]))
>>> m.sample()  # sample from a Generalized Pareto distribution with loc=0.1, scale=2.0, and concentration=0.4
tensor([ 1.5623])

Parameters:

loc (float or Tensor) – Location parameter of the distribution
scale (float or Tensor) – Scale parameter of the distribution
concentration (float or Tensor) – Concentration parameter of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Geometric

classFull reference ↗

Creates a Geometric distribution parameterized by probs, where probs is the probability of success of Bernoulli trials.

P(X=k) = (1-p)^{k} p, k = 0, 1, ...

Note

tensorplay.distributions.geometric.Geometric() $(k+1)$ -th trial is the first success hence draws samples in $\{0, 1, \ldots\}$ , whereas tensorplay.Tensor.geometric_() k-th trial is the first success hence draws samples in $\{1, 2, \ldots\}$ .

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Geometric(tensorplay.tensor([0.3]))
>>> m.sample()  # underlying Bernoulli has 30% chance 1; 70% chance 0
tensor([ 2.])

Parameters:

probs (Number, Tensor) – the probability of sampling 1. Must be in range (0, 1]
logits (Number, Tensor) – the log-odds of sampling 1.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Gumbel

classFull reference ↗

class tensorplay.distributions.Gumbel(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Samples from a Gumbel Distribution.

Examples:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Gumbel(tensorplay.tensor([1.0]), tensorplay.tensor([2.0]))
>>> m.sample()  # sample from Gumbel distribution with loc=1, scale=2
tensor([ 1.0124])

Parameters:

loc (float or Tensor) – Location parameter of the distribution
scale (float or Tensor) – Scale parameter of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

HalfCauchy

classFull reference ↗

class tensorplay.distributions.HalfCauchy(scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a half-Cauchy distribution parameterized by scale where:

X ~ Cauchy(0, scale)
Y = |X| ~ HalfCauchy(scale)

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = HalfCauchy(tensorplay.tensor([1.0]))
>>> m.sample()  # half-cauchy distributed with scale=1
tensor([ 2.3214])

Parameters:: scale (float or Tensor) – scale of the full Cauchy distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

HalfNormal

classFull reference ↗

class tensorplay.distributions.HalfNormal(scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a half-normal distribution parameterized by scale where:

X ~ Normal(0, scale)
Y = |X| ~ HalfNormal(scale)

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = HalfNormal(tensorplay.tensor([1.0]))
>>> m.sample()  # half-normal distributed with scale=1
tensor([ 0.1046])

Parameters:: scale (float or Tensor) – scale of the full Normal distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Independent

classFull reference ↗

class tensorplay.distributions.Independent(base_distribution: D, reinterpreted_batch_ndims: int, validate_args: bool | None = None)[source]

Reinterprets some of the batch dims of a distribution as event dims.

This is mainly useful for changing the shape of the result of log_prob(). For example to create a diagonal Normal distribution with the same shape as a Multivariate Normal distribution (so they are interchangeable), you can:

>>> from tensorplay.distributions.multivariate_normal import MultivariateNormal
>>> from tensorplay.distributions.normal import Normal
>>> loc = tensorplay.zeros(3)
>>> scale = tensorplay.ones(3)
>>> mvn = MultivariateNormal(loc, scale_tril=tensorplay.diag(scale))
>>> [mvn.batch_shape, mvn.event_shape]
[tensorplay.Size([]), tensorplay.Size([3])]
>>> normal = Normal(loc, scale)
>>> [normal.batch_shape, normal.event_shape]
[tensorplay.Size([3]), tensorplay.Size([])]
>>> diagn = Independent(normal, 1)
>>> [diagn.batch_shape, diagn.event_shape]
[tensorplay.Size([]), tensorplay.Size([3])]

Parameters:

base_distribution (tensorplay.distributions.distribution.Distribution) – a base distribution
reinterpreted_batch_ndims (int) – the number of batch dims to reinterpret as event dims

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

IndependentTransform

classFull reference ↗

class tensorplay.distributions.IndependentTransform(base_transform: Transform, reinterpreted_batch_ndims: int, cache_size: int = 0)[source]

Wrapper around another transform to treat reinterpreted_batch_ndims-many extra of the right most dimensions as dependent. This has no effect on the forward or backward transforms, but does sum out reinterpreted_batch_ndims-many of the rightmost dimensions in log_abs_det_jacobian().

Parameters:

base_transform (Transform) – A base transform.
reinterpreted_batch_ndims (int) – The number of extra rightmost dimensions to treat as dependent.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

InverseGamma

classFull reference ↗

class tensorplay.distributions.InverseGamma(concentration: Tensor | float, rate: Tensor | float, validate_args: bool | None = None)[source]

Creates an inverse gamma distribution parameterized by concentration and rate where:

X ~ Gamma(concentration, rate)
Y = 1 / X ~ InverseGamma(concentration, rate)

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = InverseGamma(tensorplay.tensor([2.0]), tensorplay.tensor([3.0]))
>>> m.sample()
tensor([ 1.2953])

Parameters:

concentration (float or Tensor) – shape parameter of the distribution (often referred to as alpha)
rate (float or Tensor) – rate = 1 / scale of the distribution (often referred to as beta)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Kumaraswamy

classFull reference ↗

class tensorplay.distributions.Kumaraswamy(concentration1: Tensor | float, concentration0: Tensor | float, validate_args: bool | None = None)[source]

Samples from a Kumaraswamy distribution.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Kumaraswamy(tensorplay.tensor([1.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # sample from a Kumaraswamy distribution with concentration alpha=1 and beta=1
tensor([ 0.1729])

Parameters:

concentration1 (float or Tensor) – 1st concentration parameter of the distribution (often referred to as alpha)
concentration0 (float or Tensor) – 2nd concentration parameter of the distribution (often referred to as beta)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Laplace

classFull reference ↗

class tensorplay.distributions.Laplace(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a Laplace distribution parameterized by loc and scale.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Laplace(tensorplay.tensor([0.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # Laplace distributed with loc=0, scale=1
tensor([ 0.1046])

Parameters:

loc (float or Tensor) – mean of the distribution
scale (float or Tensor) – scale of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

LKJCholesky

classFull reference ↗

class tensorplay.distributions.LKJCholesky(dim: int, concentration: Tensor | float = 1.0, validate_args: bool | None = None)[source]

LKJ distribution for lower Cholesky factor of correlation matrices. The distribution is controlled by concentration parameter $\eta$ to make the probability of the correlation matrix $M$ generated from a Cholesky factor proportional to $\det(M)^{\eta - 1}$ . Because of that, when concentration == 1, we have a uniform distribution over Cholesky factors of correlation matrices:

L ~ LKJCholesky(dim, concentration)
X = L @ L' ~ LKJCorr(dim, concentration)

Note that this distribution samples the Cholesky factor of correlation matrices and not the correlation matrices themselves and thereby differs slightly from the derivations in [1] for the LKJCorr distribution. For sampling, this uses the Onion method from [1] Section 3.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> l = LKJCholesky(3, 0.5)
>>> l.sample()  # l @ l.T is a sample of a correlation 3x3 matrix
tensor([[ 1.0000,  0.0000,  0.0000],
        [ 0.3516,  0.9361,  0.0000],
        [-0.1899,  0.4748,  0.8593]])

Parameters:

dimension (dim) – dimension of the matrices
concentration (float or Tensor) – concentration/shape parameter of the distribution (often referred to as eta)

References

[1] Generating random correlation matrices based on vines and extended onion method (2009), Daniel Lewandowski, Dorota Kurowicka, Harry Joe. Journal of Multivariate Analysis. 100. 10.1016/j.jmva.2009.04.008

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: Returns the variance of the distribution.

LogisticNormal

classFull reference ↗

class tensorplay.distributions.LogisticNormal(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a logistic-normal distribution parameterized by loc and scale that define the base Normal distribution transformed with the StickBreakingTransform such that:

X ~ LogisticNormal(loc, scale)
Y = log(X / (1 - X.cumsum(-1)))[..., :-1] ~ Normal(loc, scale)

Parameters:

loc (float or Tensor) – mean of the base distribution
scale (float or Tensor) – standard deviation of the base distribution

Example:

>>> # logistic-normal distributed with mean=(0, 0, 0) and stddev=(1, 1, 1)
>>> # of the base Normal distribution
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = LogisticNormal(tensorplay.tensor([0.0] * 3), tensorplay.tensor([1.0] * 3))
>>> m.sample()
tensor([ 0.7653,  0.0341,  0.0579,  0.1427])

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: Returns the variance of the distribution.

LogNormal

classFull reference ↗

class tensorplay.distributions.LogNormal(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a log-normal distribution parameterized by loc and scale where:

X ~ Normal(loc, scale)
Y = exp(X) ~ LogNormal(loc, scale)

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = LogNormal(tensorplay.tensor([0.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # log-normal distributed with mean=0 and stddev=1
tensor([ 0.1046])

Parameters:

loc (float or Tensor) – mean of log of distribution
scale (float or Tensor) – standard deviation of log of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

LowerCholeskyTransform

classFull reference ↗

class tensorplay.distributions.LowerCholeskyTransform(cache_size: int = 0)[source]

Transform from unconstrained matrices to lower-triangular matrices with nonnegative diagonal entries.

This is useful for parameterizing positive definite matrices in terms of their Cholesky factorization.

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

log_abs_det_jacobian(x, y): Computes the log det jacobian log |dy/dx| given input and output.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

LowRankMultivariateNormal

classFull reference ↗

class tensorplay.distributions.LowRankMultivariateNormal(loc: Tensor, cov_factor: Tensor, cov_diag: Tensor, validate_args: bool | None = None)[source]

Creates a multivariate normal distribution with covariance matrix having a low-rank form parameterized by cov_factor and cov_diag:

covariance_matrix = cov_factor @ cov_factor.T + cov_diag

Example

>>> # xdoctest: +REQUIRES(env:TENSORPLAY_DOCTEST_LAPACK)
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = LowRankMultivariateNormal(
...     tensorplay.zeros(2), tensorplay.tensor([[1.0], [0.0]]), tensorplay.ones(2)
... )
>>> m.sample()  # normally distributed with mean=`[0,0]`, cov_factor=`[[1],[0]]`, cov_diag=`[1,1]`
tensor([-0.2102, -0.5429])

Parameters:

loc (Tensor) – mean of the distribution with shape batch_shape + event_shape
cov_factor (Tensor) – factor part of low-rank form of covariance matrix with shape batch_shape + event_shape + (rank,)
cov_diag (Tensor) – diagonal part of low-rank form of covariance matrix with shape batch_shape + event_shape

Note

The computation for determinant and inverse of covariance matrix is avoided when cov_factor.shape[1] << cov_factor.shape[0] thanks to Woodbury matrix identity and matrix determinant lemma. Thanks to these formulas, we just need to compute the determinant and inverse of the small size “capacitance” matrix:

capacitance = I + cov_factor.T @ inv(cov_diag) @ cov_factor

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

MixtureSameFamily

classFull reference ↗

class tensorplay.distributions.MixtureSameFamily(mixture_distribution: Categorical, component_distribution: Distribution, validate_args: bool | None = None)[source]

The MixtureSameFamily distribution implements a (batch of) mixture distribution where all components are from different parameterizations of the same distribution type. It is parameterized by a Categorical “selecting distribution” (over k components) and a component distribution, i.e., a Distribution with a rightmost batch shape (equal to [k]) which indexes each (batch of) component.

Examples:

>>> # xdoctest: +SKIP("undefined vars")
>>> # Construct Gaussian Mixture Model in 1D consisting of 5 equally
>>> # weighted normal distributions
>>> mix = D.Categorical(tensorplay.ones(5,))
>>> comp = D.Normal(tensorplay.randn(5,), tensorplay.rand(5,))
>>> gmm = MixtureSameFamily(mix, comp)

>>> # Construct Gaussian Mixture Model in 2D consisting of 5 equally
>>> # weighted bivariate normal distributions
>>> mix = D.Categorical(tensorplay.ones(5,))
>>> comp = D.Independent(D.Normal(
...          tensorplay.randn(5,2), tensorplay.rand(5,2)), 1)
>>> gmm = MixtureSameFamily(mix, comp)

>>> # Construct a batch of 3 Gaussian Mixture Models in 2D each
>>> # consisting of 5 random weighted bivariate normal distributions
>>> mix = D.Categorical(tensorplay.rand(3,5))
>>> comp = D.Independent(D.Normal(
...         tensorplay.randn(3,5,2), tensorplay.rand(3,5,2)), 1)
>>> gmm = MixtureSameFamily(mix, comp)

Parameters:

mixture_distribution – tensorplay.distributions.Categorical-like instance. Manages the probability of selecting components. The number of categories must match the rightmost batch dimension of the component_distribution. Must have either scalar batch_shape or batch_shape matching component_distribution.batch_shape[:-1]
component_distribution – tensorplay.distributions.Distribution-like instance. Right-most batch dimension indexes component.

property batch_shape: Size: Returns the shape over which parameters are batched.

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Multinomial

classFull reference ↗

class tensorplay.distributions.Multinomial(total_count: int = 1, probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a Multinomial distribution parameterized by total_count and either probs or logits (but not both). The innermost dimension of probs indexes over categories. All other dimensions index over batches.

Note that total_count need not be specified if only log_prob() is called (see example below)

Note

sample() requires a single shared total_count for all parameters and samples.
log_prob() allows different total_count for each parameter and sample.

Example:

>>> # xdoctest: +SKIP("FIXME: found invalid values")
>>> m = Multinomial(100, tensorplay.tensor([ 1., 1., 1., 1.]))
>>> x = m.sample()  # equal probability of 0, 1, 2, 3
tensor([ 21.,  24.,  30.,  25.])

>>> Multinomial(probs=tensorplay.tensor([1., 1., 1., 1.])).log_prob(x)
tensor([-4.1338])

Parameters:

total_count (int) – number of trials
probs (Tensor) – event probabilities
logits (Tensor) – event log probabilities (unnormalized)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

MultivariateNormal

classFull reference ↗

class tensorplay.distributions.MultivariateNormal(loc: Tensor, covariance_matrix: Tensor | None = None, precision_matrix: Tensor | None = None, scale_tril: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a multivariate normal (also called Gaussian) distribution parameterized by a mean vector and a covariance matrix.

The multivariate normal distribution can be parameterized either in terms of a positive definite covariance matrix $\mathbf{\Sigma}$ or a positive definite precision matrix $\mathbf{\Sigma}^{-1}$ or a lower-triangular matrix $\mathbf{L}$ with positive-valued diagonal entries, such that $\mathbf{\Sigma} = \mathbf{L}\mathbf{L}^\top$ . This triangular matrix can be obtained via e.g. Cholesky decomposition of the covariance.

Example

>>> # xdoctest: +REQUIRES(env:TENSORPLAY_DOCTEST_LAPACK)
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = MultivariateNormal(tensorplay.zeros(2), tensorplay.eye(2))
>>> m.sample()  # normally distributed with mean=`[0,0]` and covariance_matrix=`I`
tensor([-0.2102, -0.5429])

Parameters:

loc (Tensor) – mean of the distribution
covariance_matrix (Tensor) – positive-definite covariance matrix
precision_matrix (Tensor) – positive-definite precision matrix
scale_tril (Tensor) – lower-triangular factor of covariance, with positive-valued diagonal

Note

Only one of covariance_matrix or precision_matrix or scale_tril can be specified.

Using scale_tril will be more efficient: all computations internally are based on scale_tril. If covariance_matrix or precision_matrix is passed instead, it is only used to compute the corresponding lower triangular matrices using a Cholesky decomposition.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

NegativeBinomial

classFull reference ↗

class tensorplay.distributions.NegativeBinomial(total_count: Tensor | float, probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a Negative Binomial distribution, i.e. distribution of the number of successful independent and identical Bernoulli trials before total_count failures are achieved. The probability of success of each Bernoulli trial is probs.

Parameters:

total_count (float or Tensor) – non-negative number of negative Bernoulli trials to stop, although the distribution is still valid for real valued count
probs (Tensor) – Event probabilities of success in the half open interval [0, 1)
logits (Tensor) – Event log-odds for probabilities of success

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Normal

classFull reference ↗

class tensorplay.distributions.Normal(loc: Tensor | float, scale: Tensor | float, validate_args: bool | None = None)[source]

Creates a normal (also called Gaussian) distribution parameterized by loc and scale.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Normal(tensorplay.tensor([0.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # normally distributed with loc=0 and scale=1
tensor([ 0.1046])

Parameters:

loc (float or Tensor) – mean of the distribution (often referred to as mu)
scale (float or Tensor) – standard deviation of the distribution (often referred to as sigma)

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

OneHotCategorical

classFull reference ↗

class tensorplay.distributions.OneHotCategorical(probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a one-hot categorical distribution parameterized by probs or logits.

Samples are one-hot coded vectors of size probs.size(-1).

Note

See also: tensorplay.distributions.Categorical() for specifications of probs and logits.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = OneHotCategorical(tensorplay.tensor([ 0.25, 0.25, 0.25, 0.25 ]))
>>> m.sample()  # equal probability of 0, 1, 2, 3
tensor([ 0.,  0.,  0.,  1.])

Parameters:

probs (Tensor) – event probabilities
logits (Tensor) – event log probabilities (unnormalized)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

OneHotCategoricalStraightThrough

classFull reference ↗

class tensorplay.distributions.OneHotCategoricalStraightThrough(probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a reparameterizable OneHotCategorical distribution based on the straight- through gradient estimator from [1].

[1] Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation (Bengio et al., 2013)

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Pareto

classFull reference ↗

class tensorplay.distributions.Pareto(scale: Tensor | float, alpha: Tensor | float, validate_args: bool | None = None)[source]

Samples from a Pareto Type 1 distribution.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Pareto(tensorplay.tensor([1.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # sample from a Pareto distribution with scale=1 and alpha=1
tensor([ 1.5623])

Parameters:

scale (float or Tensor) – Scale parameter of the distribution
alpha (float or Tensor) – Shape parameter of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Poisson

classFull reference ↗

class tensorplay.distributions.Poisson(rate: Tensor | bool | int | float, validate_args: bool | None = None)[source]

Creates a Poisson distribution parameterized by rate, the rate parameter.

Samples are nonnegative integers, with a pmf given by

\mathrm{rate}^k \frac{e^{-\mathrm{rate}}}{k!}

Example:

>>> # xdoctest: +SKIP("poisson_cpu not implemented for 'Long'")
>>> m = Poisson(tensorplay.tensor([4]))
>>> m.sample()
tensor([ 3.])

Parameters:: rate (Number, Tensor) – the rate parameter

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy(): Method to compute the entropy using Bregman divergence of the log normalizer.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

PositiveDefiniteTransform

classFull reference ↗

class tensorplay.distributions.PositiveDefiniteTransform(cache_size: int = 0)[source]

Transform from unconstrained matrices to positive-definite matrices.

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

log_abs_det_jacobian(x, y): Computes the log det jacobian log |dy/dx| given input and output.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

PowerTransform

classFull reference ↗

class tensorplay.distributions.PowerTransform(exponent: Tensor, cache_size: int = 0)[source]

Transform via the mapping $y = x^{\text{exponent}}$ .

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

RelaxedBernoulli

classFull reference ↗

Creates a RelaxedBernoulli distribution, parameterized by temperature, and either probs or logits (but not both). This is a relaxed version of the Bernoulli distribution, so the values are in (0, 1), and has reparametrizable samples.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = RelaxedBernoulli(tensorplay.tensor([2.2]),
...                      tensorplay.tensor([0.1, 0.2, 0.3, 0.99]))
>>> m.sample()
tensor([ 0.2951,  0.3442,  0.8918,  0.9021])

Parameters:

temperature (Tensor) – relaxation temperature
probs (Number, Tensor) – the probability of sampling 1
logits (Number, Tensor) – the log-odds of sampling 1

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: Returns the variance of the distribution.

RelaxedOneHotCategorical

classFull reference ↗

class tensorplay.distributions.RelaxedOneHotCategorical(temperature: Tensor, probs: Tensor | None = None, logits: Tensor | None = None, validate_args: bool | None = None)[source]

Creates a RelaxedOneHotCategorical distribution parameterized by temperature, and either probs or logits. This is a relaxed version of the OneHotCategorical distribution, so its samples are on simplex, and are reparametrizable.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = RelaxedOneHotCategorical(tensorplay.tensor([2.2]),
...                              tensorplay.tensor([0.1, 0.2, 0.3, 0.4]))
>>> m.sample()
tensor([ 0.1294,  0.2324,  0.3859,  0.2523])

Parameters:

temperature (Tensor) – relaxation temperature
probs (Tensor) – event probabilities
logits (Tensor) – unnormalized log probability for each event

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: Returns the variance of the distribution.

ReshapeTransform

classFull reference ↗

class tensorplay.distributions.ReshapeTransform(in_shape: Size, out_shape: Size, cache_size: int = 0)[source]

Unit Jacobian transform to reshape the rightmost part of a tensor.

Note that in_shape and out_shape must have the same number of elements, just as for tensorplay.Tensor.reshape().

Parameters:

in_shape (tensorplay.Size) – The input event shape.
out_shape (tensorplay.Size) – The output event shape.
cache_size (int) – Size of cache. If zero, no caching is done. If one, the latest single value is cached. Only 0 and 1 are supported. (Default 0.)

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

SigmoidTransform

classFull reference ↗

class tensorplay.distributions.SigmoidTransform(cache_size: int = 0)[source]

Transform via the mapping $y = \frac{1}{1 + \exp(-x)}$ and $x = \text{logit}(y)$ .

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

SoftmaxTransform

classFull reference ↗

class tensorplay.distributions.SoftmaxTransform(cache_size: int = 0)[source]

Transform from unconstrained space to the simplex via $y = \exp(x)$ then normalizing.

This is not bijective and cannot be used for HMC. However this acts mostly coordinate-wise (except for the final normalization), and thus is appropriate for coordinate-wise optimization algorithms.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

log_abs_det_jacobian(x, y): Computes the log det jacobian log |dy/dx| given input and output.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

SoftplusTransform

classFull reference ↗

class tensorplay.distributions.SoftplusTransform(cache_size: int = 0)[source]

Transform via the mapping $\text{Softplus}(x) = \log(1 + \exp(x))$ . The implementation reverts to the linear function when $x > 20$ .

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

StackTransform

classFull reference ↗

class tensorplay.distributions.StackTransform(tseq: Sequence[Transform], dim: int = 0, cache_size: int = 0)[source]

Transform functor that applies a sequence of transforms tseq component-wise to each submatrix at dim in a way compatible with tensorplay.stack().

Example:

x = tensorplay.stack([tensorplay.range(1, 10), tensorplay.range(1, 10)], dim=1)
t = StackTransform([ExpTransform(), identity_transform], dim=1)
y = t(x)

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

StickBreakingTransform

classFull reference ↗

class tensorplay.distributions.StickBreakingTransform(cache_size: int = 0)[source]

Transform from unconstrained space to the simplex of one additional dimension via a stick-breaking process.

This transform arises as an iterated sigmoid transform in a stick-breaking construction of the Dirichlet distribution: the first logit is transformed via sigmoid to the first probability and the probability of everything else, and then the process recurses.

This is bijective and appropriate for use in HMC; however it mixes coordinates together and is less appropriate for optimization.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

StudentT

classFull reference ↗

class tensorplay.distributions.StudentT(df: Tensor | float, loc: Tensor | float = 0.0, scale: Tensor | float = 1.0, validate_args: bool | None = None)[source]

Creates a Student’s t-distribution parameterized by degree of freedom df, mean loc and scale scale.

Example:

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = StudentT(tensorplay.tensor([2.0]))
>>> m.sample()  # Student's t-distributed with degrees of freedom=2
tensor([ 0.1046])

Parameters:

df (float or Tensor) – degrees of freedom
loc (float or Tensor) – mean of the distribution
scale (float or Tensor) – scale of the distribution

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

TanhTransform

classFull reference ↗

class tensorplay.distributions.TanhTransform(cache_size: int = 0)[source]

Transform via the mapping $y = \tanh(x)$ .

It is equivalent to

ComposeTransform(
    [
        AffineTransform(0.0, 2.0),
        SigmoidTransform(),
        AffineTransform(-1.0, 2.0),
    ]
)

However this might not be numerically stable, thus it is recommended to use TanhTransform instead.

Note that one should use cache_size=1 when it comes to NaN/Inf values.

forward_shape(shape): Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape): Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

Transform

classFull reference ↗

class tensorplay.distributions.Transform(cache_size: int = 0)[source]

Abstract class for invertible transformations with computable log det jacobians. They are primarily used in tensorplay.distributions.TransformedDistribution.

Caching is useful for transforms whose inverses are either expensive or numerically unstable. Note that care must be taken with memoized values since the autograd graph may be reversed. For example while the following works with or without caching:

y = t(x)
t.log_abs_det_jacobian(x, y).backward()  # x will receive gradients.

However the following will error when caching due to dependency reversal:

y = t(x)
z = t.inv(y)
grad(z.sum(), [y])  # error because z is x

Derived classes should implement one or both of _call() or _inverse(). Derived classes that set bijective=True should also implement log_abs_det_jacobian().

Parameters:

cache_size (int) – Size of cache. If zero, no caching is done. If one, the latest single value is cached. Only 0 and 1 are supported.

Variables:

domain (Constraint) – The constraint representing valid inputs to this transform.
codomain (Constraint) – The constraint representing valid outputs to this transform which are inputs to the inverse transform.
bijective (bool) – Whether this transform is bijective. A transform t is bijective iff t.inv(t(x)) == x and t(t.inv(y)) == y for every x in the domain and y in the codomain. Transforms that are not bijective should at least maintain the weaker pseudoinverse properties t(t.inv(t(x)) == t(x) and t.inv(t(t.inv(y))) == t.inv(y).
sign (int or Tensor) – For bijective univariate transforms, this should be +1 or -1 depending on whether transform is monotone increasing or decreasing.

forward_shape(shape)[source]: Infers the shape of the forward computation, given the input shape. Defaults to preserving shape.

property inv: Transform: Returns the inverse Transform of this transform. This should satisfy t.inv.inv is t.

inverse_shape(shape)[source]: Infers the shapes of the inverse computation, given the output shape. Defaults to preserving shape.

log_abs_det_jacobian(x, y)[source]: Computes the log det jacobian log |dy/dx| given input and output.

property sign: int: Returns the sign of the determinant of the Jacobian, if applicable. In general this only makes sense for bijective transforms.

TransformedDistribution

classFull reference ↗

class tensorplay.distributions.TransformedDistribution(base_distribution: Distribution, transforms: Transform | list[Transform], validate_args: bool | None = None)[source]

Extension of the Distribution class, which applies a sequence of Transforms to a base distribution. Let f be the composition of transforms applied:

X ~ BaseDistribution
Y = f(X) ~ TransformedDistribution(BaseDistribution, f)
log p(Y) = log p(X) + log |det (dX/dY)|

Note that the .event_shape of a TransformedDistribution is the maximum shape of its base distribution and its transforms, since transforms can introduce correlations among events.

An example for the usage of TransformedDistribution would be:

# Building a Logistic Distribution
# X ~ Uniform(0, 1)
# f = a + b * logit(X)
# Y ~ f(X) ~ Logistic(a, b)
base_distribution = Uniform(0, 1)
transforms = [SigmoidTransform().inv, AffineTransform(loc=a, scale=b)]
logistic = TransformedDistribution(base_distribution, transforms)

For more examples, please look at the implementations of Gumbel, HalfCauchy, HalfNormal, LogNormal, Pareto, Weibull, RelaxedBernoulli and RelaxedOneHotCategorical

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value)[source]: Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value)[source]: Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value)[source]: Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

property mean: Tensor: Returns the mean of the distribution.

property mode: Tensor: Returns the mode of the distribution.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor[source]: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=())[source]: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: Returns the variance of the distribution.

Uniform

classFull reference ↗

class tensorplay.distributions.Uniform(low: Tensor | float, high: Tensor | float, validate_args: bool | None = None)[source]

Generates uniformly distributed random samples from the half-open interval [low, high).

Example:

>>> m = Uniform(tensorplay.tensor([0.0]), tensorplay.tensor([5.0]))
>>> m.sample()  # uniformly distributed in the range [0.0, 5.0)
>>> # xdoctest: +SKIP
tensor([ 2.3418])

Parameters:

low (float or Tensor) – lower range (inclusive).
high (float or Tensor) – upper range (exclusive).

property batch_shape: Size: Returns the shape over which parameters are batched.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

VonMises

classFull reference ↗

class tensorplay.distributions.VonMises(loc: Tensor, concentration: Tensor, validate_args: bool | None = None)[source]

A circular von Mises distribution.

This implementation uses polar coordinates. The loc and value args can be any real number (to facilitate unconstrained optimization), but are interpreted as angles modulo 2 pi.

Example::

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = VonMises(tensorplay.tensor([1.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # von Mises distributed with loc=1 and concentration=1
tensor([1.9777])

Parameters:

loc (tensorplay.Tensor) – an angle in radians.
concentration (tensorplay.Tensor) – concentration parameter

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

entropy() → Tensor

Returns entropy of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

property mean: Tensor: The provided mean is the circular one.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched.

sample(sample_shape=())[source]

The sampling algorithm for the von Mises distribution is based on the following paper: D.J. Best and N.I. Fisher, “Efficient simulation of the von Mises distribution.” Applied Statistics (1979): 152-157.

Sampling is always done in double precision internally to avoid a hang in _rejection_sample() for small values of the concentration, which starts to happen for single precision around 1e-4 (see issue #88443).

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

property variance: Tensor: The provided variance is the circular one.

Weibull

classFull reference ↗

class tensorplay.distributions.Weibull(scale: Tensor | float, concentration: Tensor | float, validate_args: bool | None = None)[source]

Samples from a two-parameter Weibull distribution.

Example

>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> m = Weibull(tensorplay.tensor([1.0]), tensorplay.tensor([1.0]))
>>> m.sample()  # sample from a Weibull distribution with scale=1, concentration=1
tensor([ 0.4784])

Parameters:

scale (float or Tensor) – Scale parameter of distribution (lambda).
concentration (float or Tensor) – Concentration parameter of distribution (k/shape).
validate_args (bool, optional) – Whether to validate arguments. Default: None.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value): Computes the cumulative distribution function by inverting the transform(s) and computing the score of the base distribution.

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value): Computes the inverse cumulative distribution function using transform(s) and computing the score of the base distribution.

log_prob(value): Scores the sample by inverting the transform(s) and computing the score using the score of the base distribution and the log abs det jacobian.

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample(sample_shape=()): Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. Samples first from base distribution and applies transform() for every transform in the list.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

Wishart

classFull reference ↗

Creates a Wishart distribution parameterized by a symmetric positive definite matrix $\Sigma$ , or its Cholesky decomposition $\mathbf{\Sigma} = \mathbf{L}\mathbf{L}^\top$

Example

>>> # xdoctest: +SKIP("FIXME: scale_tril must be at least two-dimensional")
>>> m = Wishart(tensorplay.Tensor([2]), covariance_matrix=tensorplay.eye(2))
>>> m.sample()  # Wishart distributed with mean=`df * I` and
>>> # variance(x_ij)=`df` for i != j and variance(x_ij)=`2 * df` for i == j

Parameters:

df (float or Tensor) – real-valued parameter larger than the (dimension of Square matrix) - 1
covariance_matrix (Tensor) – positive-definite covariance matrix
precision_matrix (Tensor) – positive-definite precision matrix
scale_tril (Tensor) – lower-triangular factor of covariance, with positive-valued diagonal

Note

Only one of covariance_matrix or precision_matrix or scale_tril can be specified. Using scale_tril will be more efficient: all computations internally are based on scale_tril. If covariance_matrix or precision_matrix is passed instead, it is only used to compute the corresponding lower triangular matrices using a Cholesky decomposition. ‘tensorplay.distributions.LKJCholesky’ is a restricted Wishart distribution.[1]

References

[1] Wang, Z., Wu, Y. and Chu, H., 2018. On equivalence of the LKJ distribution and the restricted Wishart distribution. [2] Sawyer, S., 2007. Wishart Distributions and Inverse-Wishart Sampling. [3] Anderson, T. W., 2003. An Introduction to Multivariate Statistical Analysis (3rd ed.). [4] Odell, P. L. & Feiveson, A. H., 1966. A Numerical Procedure to Generate a Sample Covariance Matrix. JASA, 61(313):199-203. [5] Ku, Y.-C. & Bloomfield, P., 2010. Generating Random Wishart Matrices with Fractional Degrees of Freedom in OX.

property batch_shape: Size: Returns the shape over which parameters are batched.

cdf(value: Tensor) → Tensor

Returns the cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

enumerate_support(expand: bool = True) → Tensor

To iterate over the full Cartesian product use itertools.product(m.enumerate_support()).

Parameters:: expand (bool) – whether to expand the support over the batch dims to match the distribution’s batch_shape.
Returns:: Tensor iterating over dimension 0.

property event_shape: Size: Returns the shape of a single sample (without batching).

icdf(value: Tensor) → Tensor

Returns the inverse cumulative density/mass function evaluated at value.

Parameters:: value (Tensor)

perplexity() → Tensor

Returns perplexity of distribution, batched over batch_shape.

Returns:: Tensor of shape batch_shape.

rsample(sample_shape: Size | Sequence[int] = (), max_try_correction=None) → Tensor[source]: Warning

In some cases, sampling algorithm based on Bartlett decomposition may return singular matrix samples. Several tries to correct singular samples are performed by default, but it may end up returning singular matrix samples. Singular samples may return -inf values in .log_prob(). In those cases, the user should validate the samples and either fix the value of df or adjust max_try_correction value for argument in .rsample accordingly.

sample(sample_shape: Size | Sequence[int] = ()) → Tensor: Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched.

sample_n(n: int) → Tensor: Generates n samples or n batches of samples if the distribution parameters are batched.

static set_default_validate_args(value: bool) → None

Sets whether validation is enabled or disabled.

Parameters:: value (bool) – Whether to enable validation.

property stddev: Tensor: Returns the standard deviation of the distribution.

tensorplay.distributions API

Functions 2

Classes 62

Help improve this page

On this page