Miscellaneous Ops¶
The pyro.ops
module implements tensor utilities
that are mostly independent of the rest of Pyro.
Utilities for HMC¶
- class DualAveraging(prox_center=0, t0=10, kappa=0.75, gamma=0.05)[source]¶
Bases:
object
Dual Averaging is a scheme to solve convex optimization problems. It belongs to a class of subgradient methods which uses subgradients to update parameters (in primal space) of a model. Under some conditions, the averages of generated parameters during the scheme are guaranteed to converge to an optimal value. However, a counter-intuitive aspect of traditional subgradient methods is “new subgradients enter the model with decreasing weights” (see \([1]\)). Dual Averaging scheme solves that phenomenon by updating parameters using weights equally for subgradients (which lie in a dual space), hence we have the name “dual averaging”.
This class implements a dual averaging scheme which is adapted for Markov chain Monte Carlo (MCMC) algorithms. To be more precise, we will replace subgradients by some statistics calculated during an MCMC trajectory. In addition, introducing some free parameters such as
t0
andkappa
is helpful and still guarantees the convergence of the scheme.References
[1] Primal-dual subgradient methods for convex problems, Yurii Nesterov
[2] The No-U-turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo, Matthew D. Hoffman, Andrew Gelman
- Parameters
prox_center (float) – A “prox-center” parameter introduced in \([1]\) which pulls the primal sequence towards it.
t0 (float) – A free parameter introduced in \([2]\) that stabilizes the initial steps of the scheme.
kappa (float) – A free parameter introduced in \([2]\) that controls the weights of steps of the scheme. For a small
kappa
, the scheme will quickly forget states from early steps. This should be a number in \((0.5, 1]\).gamma (float) – A free parameter which controls the speed of the convergence of the scheme.
- velocity_verlet(z, r, potential_fn, kinetic_grad, step_size, num_steps=1, z_grads=None)[source]¶
Second order symplectic integrator that uses the velocity verlet algorithm.
- Parameters
z (dict) – dictionary of sample site names and their current values (type
Tensor
).r (dict) – dictionary of sample site names and corresponding momenta (type
Tensor
).potential_fn (callable) – function that returns potential energy given z for each sample site. The negative gradient of the function with respect to
z
determines the rate of change of the corresponding sites’ momentar
.kinetic_grad (callable) – a function calculating gradient of kinetic energy w.r.t. momentum variable.
step_size (float) – step size for each time step iteration.
num_steps (int) – number of discrete time steps over which to integrate.
z_grads (torch.Tensor) – optional gradients of potential energy at current
z
.
- Return tuple (z_next, r_next, z_grads, potential_energy)
next position and momenta, together with the potential energy and its gradient w.r.t.
z_next
.
- potential_grad(potential_fn, z)[source]¶
Gradient of potential_fn w.r.t. parameters z.
- Parameters
potential_fn – python callable that takes in a dictionary of parameters and returns the potential energy.
z (dict) – dictionary of parameter values keyed by site name.
- Returns
tuple of (z_grads, potential_energy), where z_grads is a dictionary with the same keys as z containing gradients and potential_energy is a torch scalar.
- register_exception_handler(name: str, handler: Callable[[Exception], bool], warn_on_overwrite: bool = True) None [source]¶
Register an exception handler for handling (primarily numerical) errors when evaluating the potential function.
- Parameters
name – name of the handler (must be unique).
handler – A callable mapping an Exception to a boolean. Exceptions that evaluate to true in any of the handlers are handled in the computation of the potential energy.
warn_on_overwrite – If True, warns when overwriting a handler already registered under the provided name.
- class WelfordCovariance(diagonal=True)[source]¶
Bases:
object
Implements Welford’s online scheme for estimating (co)variance (see \([1]\)). Useful for adapting diagonal and dense mass structures for HMC.
References
[1] The Art of Computer Programming, Donald E. Knuth
- class WelfordArrowheadCovariance(head_size=0)[source]¶
Bases:
object
Likes
WelfordCovariance
but generalized to the arrowhead structure.
Newton Optimizers¶
- newton_step(loss, x, trust_radius=None)[source]¶
Performs a Newton update step to minimize loss on a batch of variables, optionally constraining to a trust region [1].
This is especially usful because the final solution of newton iteration is differentiable wrt the inputs, even when all but the final
x
is detached, due to this method’s quadratic convergence [2].loss
must be twice-differentiable as a function ofx
. Ifloss
is2+d
-times differentiable, then the return value of this function isd
-times differentiable.When
loss
is interpreted as a negative log probability density, then the return valuesmode,cov
of this function can be used to construct a Laplace approximationMultivariateNormal(mode,cov)
.Warning
Take care to detach the result of this function when used in an optimization loop. If you forget to detach the result of this function during optimization, then backprop will propagate through the entire iteration process, and worse will compute two extra derivatives for each step.
Example use inside a loop:
x = torch.zeros(1000, 2) # arbitrary initial value for step in range(100): x = x.detach() # block gradients through previous steps x.requires_grad = True # ensure loss is differentiable wrt x loss = my_loss_function(x) x = newton_step(loss, x, trust_radius=1.0) # the final x is still differentiable
- [1] Yuan, Ya-xiang. Iciam. Vol. 99. 2000.
“A review of trust region algorithms for optimization.” ftp://ftp.cc.ac.cn/pub/yyx/papers/p995.pdf
- [2] Christianson, Bruce. Optimization Methods and Software 3.4 (1994)
“Reverse accumulation and attractive fixed points.” http://uhra.herts.ac.uk/bitstream/handle/2299/4338/903839.pdf
- Parameters
loss (torch.Tensor) – A scalar function of
x
to be minimized.x (torch.Tensor) – A dependent variable of shape
(N, D)
whereN
is the batch size andD
is a small number.trust_radius (float) – An optional trust region trust_radius. The updated value
mode
of this function will be withintrust_radius
of the inputx
.
- Returns
A pair
(mode, cov)
wheremode
is an updated tensor of the same shape as the original valuex
, andcov
is an esitmate of the covariance DxD matrix withcov.shape == x.shape[:-1] + (D,D)
.- Return type
- newton_step_1d(loss, x, trust_radius=None)[source]¶
Performs a Newton update step to minimize loss on a batch of 1-dimensional variables, optionally regularizing to constrain to a trust region.
See
newton_step()
for details.- Parameters
loss (torch.Tensor) – A scalar function of
x
to be minimized.x (torch.Tensor) – A dependent variable with rightmost size of 1.
trust_radius (float) – An optional trust region trust_radius. The updated value
mode
of this function will be withintrust_radius
of the inputx
.
- Returns
A pair
(mode, cov)
wheremode
is an updated tensor of the same shape as the original valuex
, andcov
is an esitmate of the covariance 1x1 matrix withcov.shape == x.shape[:-1] + (1,1)
.- Return type
- newton_step_2d(loss, x, trust_radius=None)[source]¶
Performs a Newton update step to minimize loss on a batch of 2-dimensional variables, optionally regularizing to constrain to a trust region.
See
newton_step()
for details.- Parameters
loss (torch.Tensor) – A scalar function of
x
to be minimized.x (torch.Tensor) – A dependent variable with rightmost size of 2.
trust_radius (float) – An optional trust region trust_radius. The updated value
mode
of this function will be withintrust_radius
of the inputx
.
- Returns
A pair
(mode, cov)
wheremode
is an updated tensor of the same shape as the original valuex
, andcov
is an esitmate of the covariance 2x2 matrix withcov.shape == x.shape[:-1] + (2,2)
.- Return type
- newton_step_3d(loss, x, trust_radius=None)[source]¶
Performs a Newton update step to minimize loss on a batch of 3-dimensional variables, optionally regularizing to constrain to a trust region.
See
newton_step()
for details.- Parameters
loss (torch.Tensor) – A scalar function of
x
to be minimized.x (torch.Tensor) – A dependent variable with rightmost size of 2.
trust_radius (float) – An optional trust region trust_radius. The updated value
mode
of this function will be withintrust_radius
of the inputx
.
- Returns
A pair
(mode, cov)
wheremode
is an updated tensor of the same shape as the original valuex
, andcov
is an esitmate of the covariance 3x3 matrix withcov.shape == x.shape[:-1] + (3,3)
.- Return type
Special Functions¶
- safe_log(x)[source]¶
Like
torch.log()
but avoids infinite gradients at log(0) by clamping them to at most1 / finfo.eps
.
- log_beta(x, y, tol=0.0)[source]¶
Computes log Beta function.
When
tol >= 0.02
this uses a shifted Stirling’s approximation to the log Beta function. The approximation adapts Stirling’s approximation of the log Gamma function:lgamma(z) ≈ (z - 1/2) * log(z) - z + log(2 * pi) / 2
to approximate the log Beta function:
log_beta(x, y) ≈ ((x-1/2) * log(x) + (y-1/2) * log(y) - (x+y-1/2) * log(x+y) + log(2*pi)/2)
The approximation additionally improves accuracy near zero by iteratively shifting the log Gamma approximation using the recursion:
lgamma(x) = lgamma(x + 1) - log(x)
If this recursion is applied
n
times, then absolute error is bounded byerror < 0.082 / n < tol
, thus we choosen
based on the user providedtol
.- Parameters
x (torch.Tensor) – A positive tensor.
y (torch.Tensor) – A positive tensor.
tol (float) – Bound on maximum absolute error. Defaults to 0.1. For very small
tol
, this function simply defers tolog_beta()
.
- Return type
- log_binomial(n, k, tol=0.0)[source]¶
Computes log binomial coefficient.
When
tol >= 0.02
this uses a shifted Stirling’s approximation to the log Beta function vialog_beta()
.- Parameters
n (torch.Tensor) – A nonnegative integer tensor.
k (torch.Tensor) – An integer tensor ranging in
[0, n]
.
- Return type
- log_I1(orders: int, value: torch.Tensor, terms=250)[source]¶
Compute first n log modified bessel function of first kind .. math
\log(I_v(z)) = v*\log(z/2) + \log(\sum_{k=0}^\inf \exp\left[2*k*\log(z/2) - \sum_kk^k log(kk) - \lgamma(v + k + 1)\right])
- Parameters
orders – orders of the log modified bessel function.
value – values to compute modified bessel function for
terms – truncation of summation
- Returns
0 to orders modified bessel function
- get_quad_rule(num_quad, prototype_tensor)[source]¶
Get quadrature points and corresponding log weights for a Gauss Hermite quadrature rule with the specified number of quadrature points.
Example usage:
quad_points, log_weights = get_quad_rule(32, prototype_tensor) # transform to N(0, 4.0) Normal distribution quad_points *= 4.0 # compute variance integral in log-space using logsumexp and exponentiate variance = torch.logsumexp(quad_points.pow(2.0).log() + log_weights, axis=0).exp() assert (variance - 16.0).abs().item() < 1.0e-6
- Parameters
num_quad (int) – number of quadrature points.
prototype_tensor (torch.Tensor) – used to determine dtype and device of returned tensors.
- Returns
tuple of torch.Tensor`s of the form `(quad_points, log_weights)
- sparse_multinomial_likelihood(total_count, nonzero_logits, nonzero_value)[source]¶
The following are equivalent:
# Version 1. dense log_prob = Multinomial(logits=logits).log_prob(value).sum() # Version 2. sparse nnz = value.nonzero(as_tuple=True) log_prob = sparse_multinomial_likelihood( value.sum(-1), (logits - logits.logsumexp(-1))[nnz], value[nnz], )
Tensor Utilities¶
- as_complex(x)[source]¶
Similar to
torch.view_as_complex()
but copies data in case strides are not multiples of two.
- block_diag_embed(mat)[source]¶
Takes a tensor of shape (…, B, M, N) and returns a block diagonal tensor of shape (…, B x M, B x N).
- Parameters
mat (torch.Tensor) – an input tensor with 3 or more dimensions
- Returns torch.Tensor
a block diagonal tensor with dimension m.dim() - 1
- block_diagonal(mat, block_size)[source]¶
Takes a block diagonal tensor of shape (…, B x M, B x N) and returns a tensor of shape (…, B, M, N).
- Parameters
mat (torch.Tensor) – an input tensor with 2 or more dimensions
block_size (int) – the number of blocks B.
- Returns torch.Tensor
a tensor with dimension mat.dim() + 1
- periodic_repeat(tensor, size, dim)[source]¶
Repeat a
period
-sized tensor up to givensize
. For example:>>> x = torch.tensor([[1, 2, 3], [4, 5, 6]]) >>> periodic_repeat(x, 4, 0) tensor([[1, 2, 3], [4, 5, 6], [1, 2, 3], [4, 5, 6]]) >>> periodic_repeat(x, 4, 1) tensor([[1, 2, 3, 1], [4, 5, 6, 4]])
This is useful for computing static seasonality in time series models.
- Parameters
tensor (torch.Tensor) – A tensor of differences.
size (int) – Desired size of the result along dimension
dim
.dim (int) – The tensor dimension along which to repeat.
- periodic_cumsum(tensor, period, dim)[source]¶
Compute periodic cumsum along a given dimension. For example if dim=0:
for t in range(period): assert result[t] == tensor[t] for t in range(period, len(tensor)): assert result[t] == tensor[t] + result[t - period]
This is useful for computing drifting seasonality in time series models.
- Parameters
tensor (torch.Tensor) – A tensor of differences.
period (int) – The period of repetition.
dim (int) – The tensor dimension along which to accumulate.
- periodic_features(duration, max_period=None, min_period=None, **options)[source]¶
Create periodic (sin,cos) features from
max_period
down tomin_period
.This is useful in time series models where long uneven seasonality can be treated via regression. When only
max_period
is specified this generates periodic features at all length scales. When alsomin_period
is specified this generates periodic features at large length scales, but omits high frequency features. This is useful when combining regression for long seasonality with other techniques likeperiodic_repeat()
andperiodic_cumsum()
for short time scales. For example, to combine regress yearly seasonality down to the scale of one week one could setmax_period=365.25
andmin_period=7
.- Parameters
- Returns
A
(duration, 2 * ceil(max_period / min_period) - 2)
-shaped tensor of features normalized to lie in [-1,1].- Return type
- next_fast_len(size)[source]¶
Returns the next largest number
n >= size
whose prime factors are all 2, 3, or 5. These sizes are efficient for fast fourier transforms. Equivalent toscipy.fftpack.next_fast_len()
.- Parameters
size (int) – A positive number.
- Returns
A possibly larger number.
- Rtype int
- convolve(signal, kernel, mode='full')[source]¶
Computes the 1-d convolution of signal by kernel using FFTs. The two arguments should have the same rightmost dim, but may otherwise be arbitrarily broadcastable.
- Parameters
signal (torch.Tensor) – A signal to convolve.
kernel (torch.Tensor) – A convolution kernel.
mode (str) – One of: ‘full’, ‘valid’, ‘same’.
- Returns
A tensor with broadcasted shape. Letting
m = signal.size(-1)
andn = kernel.size(-1)
, the rightmost size of the result will be:m + n - 1
if mode is ‘full’;max(m, n) - min(m, n) + 1
if mode is ‘valid’; ormax(m, n)
if mode is ‘same’.- Rtype torch.Tensor
- repeated_matmul(M, n)[source]¶
Takes a batch of matrices M as input and returns the stacked result of doing the n-many matrix multiplications \(M\), \(M^2\), …, \(M^n\). Parallel cost is logarithmic in n.
- Parameters
M (torch.Tensor) – A batch of square tensors of shape (…, N, N).
n (int) – The order of the largest product \(M^n\)
- Returns torch.Tensor
A batch of square tensors of shape (n, …, N, N)
- dct(x, dim=- 1)[source]¶
Discrete cosine transform of type II, scaled to be orthonormal.
This is the inverse of
idct_ii()
, and is equivalent toscipy.fftpack.dct()
withnorm="ortho"
.- Parameters
x (Tensor) – The input signal.
dim (int) – Dimension along which to compute DCT.
- Return type
Tensor
- idct(x, dim=- 1)[source]¶
Inverse discrete cosine transform of type II, scaled to be orthonormal.
This is the inverse of
dct_ii()
, and is equivalent toscipy.fftpack.idct()
withnorm="ortho"
.- Parameters
x (Tensor) – The input signal.
dim (int) – Dimension along which to compute DCT.
- Return type
Tensor
- haar_transform(x)[source]¶
Discrete Haar transform.
Performs a Haar transform along the final dimension. This is the inverse of
inverse_haar_transform()
.- Parameters
x (Tensor) – The input signal.
- Return type
Tensor
- inverse_haar_transform(x)[source]¶
Performs an inverse Haar transform along the final dimension. This is the inverse of
haar_transform()
.- Parameters
x (Tensor) – The input signal.
- Return type
Tensor
- safe_normalize(x, *, p=2)[source]¶
Safely project a vector onto the sphere wrt the
p
-norm. This avoids the singularity at zero by mapping zero to the vector[1, 0, 0, ..., 0]
.- Parameters
x (torch.Tensor) – A vector
p (float) – The norm exponent, defaults to 2 i.e. the Euclidean norm.
- Returns
A normalized version
x / ||x||_p
.- Return type
Tensor
Tensor Indexing¶
- index(tensor, args)[source]¶
Indexing with nested tuples.
See also the convenience wrapper
Index
.This is useful for writing indexing code that is compatible with multiple interpretations, e.g. scalar evaluation, vectorized evaluation, or reshaping.
For example suppose
x
is a parameter withx.dim() == 2
and we wish to generalize the expressionx[..., t]
wheret
can be any of:a scalar
t=1
as inx[..., 1]
;a slice
t=slice(None)
equivalent tox[..., :]
; ora reshaping operation
t=(Ellipsis, None)
equivalent tox.unsqueeze(-1)
.
While naive indexing would work for the first two , the third example would result in a nested tuple
(Ellipsis, (Ellipsis, None))
. This helper flattens that nested tuple and combines consecutiveEllipsis
.- Parameters
tensor (torch.Tensor) – A tensor to be indexed.
args (tuple) – An index, as args to
__getitem__
.
- Returns
A flattened interpetation of
tensor[args]
.- Return type
- class Index(tensor)[source]¶
Bases:
object
Convenience wrapper around
index()
.The following are equivalent:
Index(x)[..., i, j, :] index(x, (Ellipsis, i, j, slice(None)))
- Parameters
tensor (torch.Tensor) – A tensor to be indexed.
- Returns
An object with a special
__getitem__()
method.
- vindex(tensor, args)[source]¶
Vectorized advanced indexing with broadcasting semantics.
See also the convenience wrapper
Vindex
.This is useful for writing indexing code that is compatible with batching and enumeration, especially for selecting mixture components with discrete random variables.
For example suppose
x
is a parameter withx.dim() == 3
and we wish to generalize the expressionx[i, :, j]
from integeri,j
to tensorsi,j
with batch dims and enum dims (but no event dims). Then we can write the generalize version usingVindex
xij = Vindex(x)[i, :, j] batch_shape = broadcast_shape(i.shape, j.shape) event_shape = (x.size(1),) assert xij.shape == batch_shape + event_shape
To handle the case when
x
may also contain batch dimensions (e.g. ifx
was sampled in a plated context as when using vectorized particles),vindex()
uses the special convention thatEllipsis
denotes batch dimensions (hence...
can appear only on the left, never in the middle or in the right). Supposex
has event dim 3. Then we can write:old_batch_shape = x.shape[:-3] old_event_shape = x.shape[-3:] xij = Vindex(x)[..., i, :, j] # The ... denotes unknown batch shape. new_batch_shape = broadcast_shape(old_batch_shape, i.shape, j.shape) new_event_shape = (x.size(1),) assert xij.shape = new_batch_shape + new_event_shape
Note that this special handling of
Ellipsis
differs from the NEP [1].Formally, this function assumes:
Each arg is either
Ellipsis
,slice(None)
, an integer, or a batchedtorch.LongTensor
(i.e. with empty event shape). This function does not support Nontrivial slices ortorch.BoolTensor
masks.Ellipsis
can only appear on the left asargs[0]
.If
args[0] is not Ellipsis
thentensor
is not batched, and its event dim is equal tolen(args)
.If
args[0] is Ellipsis
thentensor
is batched and its event dim is equal tolen(args[1:])
. Dims oftensor
to the left of the event dims are considered batch dims and will be broadcasted with dims of tensor args.
Note that if none of the args is a tensor with
.dim() > 0
, then this function behaves like standard indexing:if not any(isinstance(a, torch.Tensor) and a.dim() for a in args): assert Vindex(x)[args] == x[args]
References
- [1] https://www.numpy.org/neps/nep-0021-advanced-indexing.html
introduces
vindex
as a helper for vectorized indexing. The Pyro implementation is similar to the proposed notationx.vindex[]
except for slightly different handling ofEllipsis
.
- Parameters
tensor (torch.Tensor) – A tensor to be indexed.
args (tuple) – An index, as args to
__getitem__
.
- Returns
A nonstandard interpetation of
tensor[args]
.- Return type
- class Vindex(tensor)[source]¶
Bases:
object
Convenience wrapper around
vindex()
.The following are equivalent:
Vindex(x)[..., i, j, :] vindex(x, (Ellipsis, i, j, slice(None)))
- Parameters
tensor (torch.Tensor) – A tensor to be indexed.
- Returns
An object with a special
__getitem__()
method.
Tensor Contraction¶
- contract(equation, *operands, **kwargs)[source]¶
Wrapper around
opt_einsum.contract()
that optionally uses Pyro’s cheap optimizer and optionally caches contraction paths.- Parameters
cache_path (bool) – whether to cache the contraction path. Defaults to True.
- contract_expression(equation, *shapes, **kwargs)[source]¶
Wrapper around
opt_einsum.contract_expression()
that optionally uses Pyro’s cheap optimizer and optionally caches contraction paths.- Parameters
cache_path (bool) – whether to cache the contraction path. Defaults to True.
- einsum(equation, *operands, **kwargs)[source]¶
Generalized plated sum-product algorithm via tensor variable elimination.
This generalizes
contract()
in two ways:Multiple outputs are allowed, and intermediate results can be shared.
Inputs and outputs can be plated along symbols given in
plates
; reductions alongplates
are product reductions.
The best way to understand this function is to try the examples below, which show how
einsum()
calls can be implemented as multiple calls tocontract()
(which is generally more expensive).To illustrate multiple outputs, note that the following are equivalent:
z1, z2, z3 = einsum('ab,bc->a,b,c', x, y) # multiple outputs z1 = contract('ab,bc->a', x, y) z2 = contract('ab,bc->b', x, y) z3 = contract('ab,bc->c', x, y)
To illustrate plated inputs, note that the following are equivalent:
assert len(x) == 3 and len(y) == 3 z = einsum('ab,ai,bi->b', w, x, y, plates='i') z = contract('ab,a,a,a,b,b,b->b', w, *x, *y)
When a sum dimension a always appears with a plate dimension i, then a corresponds to a distinct symbol for each slice of a. Thus the following are equivalent:
assert len(x) == 3 and len(y) == 3 z = einsum('ai,ai->', x, y, plates='i') z = contract('a,b,c,a,b,c->', *x, *y)
When such a sum dimension appears in the output, it must be accompanied by all of its plate dimensions, e.g. the following are equivalent:
assert len(x) == 3 and len(y) == 3 z = einsum('abi,abi->bi', x, y, plates='i') z0 = contract('ab,ac,ad,ab,ac,ad->b', *x, *y) z1 = contract('ab,ac,ad,ab,ac,ad->c', *x, *y) z2 = contract('ab,ac,ad,ab,ac,ad->d', *x, *y) z = torch.stack([z0, z1, z2])
Note that each plate slice through the output is multilinear in all plate slices through all inptus, thus e.g. batch matrix multiply would be implemented without
plates
, so the following are all equivalent:xy = einsum('abc,acd->abd', x, y, plates='') xy = torch.stack([xa.mm(ya) for xa, ya in zip(x, y)]) xy = torch.bmm(x, y)
Among all valid equations, some computations are polynomial in the sizes of the input tensors and other computations are exponential in the sizes of the input tensors. This function raises
NotImplementedError
whenever the computation is exponential.- Parameters
equation (str) – An einsum equation, optionally with multiple outputs.
operands (torch.Tensor) – A collection of tensors.
plates (str) – An optional string of plate symbols.
backend (str) – An optional einsum backend, defaults to ‘torch’.
cache (dict) – An optional
shared_intermediates()
cache.modulo_total (bool) – Optionally allow einsum to arbitrarily scale each result plate, which can significantly reduce computation. This is safe to set whenever each result plate denotes a nonnormalized probability distribution whose total is not of interest.
- Returns
a tuple of tensors of requested shape, one entry per output.
- Return type
- Raises
ValueError – if tensor sizes mismatch or an output requests a plated dim without that dim’s plates.
NotImplementedError – if contraction would have cost exponential in the size of any input tensor.
Gaussian Contraction¶
- class Gaussian(log_normalizer: torch.Tensor, info_vec: torch.Tensor, precision: torch.Tensor)[source]¶
Bases:
object
Non-normalized Gaussian distribution.
This represents an arbitrary semidefinite quadratic function, which can be interpreted as a rank-deficient scaled Gaussian distribution. The precision matrix may have zero eigenvalues, thus it may be impossible to work directly with the covariance matrix.
- Parameters
log_normalizer (torch.Tensor) – a normalization constant, which is mainly used to keep track of normalization terms during contractions.
info_vec (torch.Tensor) – information vector, which is a scaled version of the mean
info_vec = precision @ mean
. We use this represention to make gaussian contraction fast and stable.precision (torch.Tensor) – precision matrix of this gaussian.
- property batch_shape¶
- expand(batch_shape) pyro.ops.gaussian.Gaussian [source]¶
- reshape(batch_shape) pyro.ops.gaussian.Gaussian [source]¶
- __getitem__(index) pyro.ops.gaussian.Gaussian [source]¶
Index into the batch_shape of a Gaussian.
- static cat(parts, dim=0) pyro.ops.gaussian.Gaussian [source]¶
Concatenate a list of Gaussians along a given batch dimension.
- event_pad(left=0, right=0) pyro.ops.gaussian.Gaussian [source]¶
Pad along event dimension.
- event_permute(perm) pyro.ops.gaussian.Gaussian [source]¶
Permute along event dimension.
- __add__(other: Union[pyro.ops.gaussian.Gaussian, int, float, torch.Tensor]) pyro.ops.gaussian.Gaussian [source]¶
Adds two Gaussians in log-density space.
- log_density(value: torch.Tensor) torch.Tensor [source]¶
Evaluate the log density of this Gaussian at a point value:
-0.5 * value.T @ precision @ value + value.T @ info_vec + log_normalizer
This is mainly used for testing.
- rsample(sample_shape=torch.Size([]), noise: Optional[torch.Tensor] = None) torch.Tensor [source]¶
Reparameterized sampler.
- condition(value: torch.Tensor) pyro.ops.gaussian.Gaussian [source]¶
Condition this Gaussian on a trailing subset of its state. This should satisfy:
g.condition(y).dim() == g.dim() - y.size(-1)
Note that since this is a non-normalized Gaussian, we include the density of
y
in the result. Thuscondition()
is similar to afunctools.partial
binding of arguments:left = x[..., :n] right = x[..., n:] g.log_density(x) == g.condition(right).log_density(left)
- left_condition(value: torch.Tensor) pyro.ops.gaussian.Gaussian [source]¶
Condition this Gaussian on a leading subset of its state. This should satisfy:
g.condition(y).dim() == g.dim() - y.size(-1)
Note that since this is a non-normalized Gaussian, we include the density of
y
in the result. Thuscondition()
is similar to afunctools.partial
binding of arguments:left = x[..., :n] right = x[..., n:] g.log_density(x) == g.left_condition(left).log_density(right)
- marginalize(left=0, right=0) pyro.ops.gaussian.Gaussian [source]¶
Marginalizing out variables on either side of the event dimension:
g.marginalize(left=n).event_logsumexp() = g.logsumexp() g.marginalize(right=n).event_logsumexp() = g.logsumexp()
and for data
x
:- g.condition(x).event_logsumexp()
= g.marginalize(left=g.dim() - x.size(-1)).log_density(x)
- event_logsumexp() torch.Tensor [source]¶
Integrates out all latent state (i.e. operating on event dimensions).
- class AffineNormal(matrix, loc, scale)[source]¶
Bases:
object
Represents a conditional diagonal normal distribution over a random variable
Y
whose mean is an affine function of a random variableX
. The likelihood ofX
is thus:AffineNormal(matrix, loc, scale).condition(y).log_density(x)
which is equivalent to:
Normal(x @ matrix + loc, scale).to_event(1).log_prob(y)
- Parameters
matrix (torch.Tensor) – A transformation from
X
toY
. Should have rightmost shape(x_dim, y_dim)
.loc (torch.Tensor) – A constant offset for
Y
’s mean. Should have rightmost shape(y_dim,)
.scale (torch.Tensor) – Standard deviation for
Y
. Should have rightmost shape(y_dim,)
.
- property batch_shape¶
- left_condition(value)[source]¶
If
value.size(-1) == x_dim
, this returns a Normal distribution withevent_dim=1
. After applying this method, the cost to draw a sample isO(y_dim)
instead ofO(y_dim ** 3)
.
- rsample(sample_shape=torch.Size([]), noise: Optional[torch.Tensor] = None) torch.Tensor [source]¶
Reparameterized sampler.
- mvn_to_gaussian(mvn)[source]¶
Convert a MultivariateNormal distribution to a Gaussian.
- Parameters
mvn (MultivariateNormal) – A multivariate normal distribution.
- Returns
An equivalent Gaussian object.
- Return type
- matrix_and_gaussian_to_gaussian(matrix: torch.Tensor, y_gaussian: pyro.ops.gaussian.Gaussian) pyro.ops.gaussian.Gaussian [source]¶
Constructs a conditional Gaussian for
p(y|x)
wherey - x @ matrix ~ y_gaussian
.- Parameters
matrix (torch.Tensor) – A right-acting transformation matrix.
y_gaussian (Gaussian) – A distribution over noise of
y - x@matrix
.
- Return type
- matrix_and_mvn_to_gaussian(matrix, mvn)[source]¶
Convert a noisy affine function to a Gaussian. The noisy affine function is defined as:
y = x @ matrix + mvn.sample()
- gaussian_tensordot(x: pyro.ops.gaussian.Gaussian, y: pyro.ops.gaussian.Gaussian, dims: int = 0) pyro.ops.gaussian.Gaussian [source]¶
Computes the integral over two gaussians:
(x @ y)(a,c) = log(integral(exp(x(a,b) + y(b,c)), b)),
where x is a gaussian over variables (a,b), y is a gaussian over variables (b,c), (a,b,c) can each be sets of zero or more variables, and dims is the size of b.
- Parameters
x – a Gaussian instance
y – a Gaussian instance
dims – number of variables to contract
- sequential_gaussian_tensordot(gaussian: pyro.ops.gaussian.Gaussian) pyro.ops.gaussian.Gaussian [source]¶
Integrates a Gaussian
x
whose rightmost batch dimension is time, computes:x[..., 0] @ x[..., 1] @ ... @ x[..., T-1]
- sequential_gaussian_filter_sample(init: pyro.ops.gaussian.Gaussian, trans: pyro.ops.gaussian.Gaussian, sample_shape: Tuple[int, ...] = (), noise: Optional[torch.Tensor] = None) torch.Tensor [source]¶
Draws a reparameterized sample from a Markov product of Gaussians via parallel-scan forward-filter backward-sample.
- Parameters
init (Gaussian) – A Gaussian representing an initial state.
trans (Gaussian) – A Gaussian representing as series of state transitions, with time as the rightmost batch dimension. This must have twice the event dim as
init
:trans.dim() == 2 * init.dim()
.sample_shape (tuple) – An optional extra shape of samples to draw.
noise (torch.Tensor) – An optional standard white noise tensor of shape
sample_shape + batch_shape + (duration, state_dim)
, whereduration = 1 + trans.batch_shape[-1]
is the number of time points to be sampled, andstate_dim = init.dim()
is the state dimension. This is useful for computing the mean (pass zeros), varying temperature (pass scaled noise), and antithetic sampling (passcat([z,-z])
).
- Returns
A reparametrized sample of shape
sample_shape + batch_shape + (duration, state_dim)
.- Return type
Statistical Utilities¶
- gelman_rubin(input, chain_dim=0, sample_dim=1)[source]¶
Computes R-hat over chains of samples. It is required that
input.size(sample_dim) >= 2
andinput.size(chain_dim) >= 2
.- Parameters
input (torch.Tensor) – the input tensor.
chain_dim (int) – the chain dimension.
sample_dim (int) – the sample dimension.
- Returns torch.Tensor
R-hat of
input
.
- split_gelman_rubin(input, chain_dim=0, sample_dim=1)[source]¶
Computes R-hat over chains of samples. It is required that
input.size(sample_dim) >= 4
.- Parameters
input (torch.Tensor) – the input tensor.
chain_dim (int) – the chain dimension.
sample_dim (int) – the sample dimension.
- Returns torch.Tensor
split R-hat of
input
.
- autocorrelation(input, dim=0)[source]¶
Computes the autocorrelation of samples at dimension
dim
.Reference: https://en.wikipedia.org/wiki/Autocorrelation#Efficient_computation
- Parameters
input (torch.Tensor) – the input tensor.
dim (int) – the dimension to calculate autocorrelation.
- Returns torch.Tensor
autocorrelation of
input
.
- autocovariance(input, dim=0)[source]¶
Computes the autocovariance of samples at dimension
dim
.- Parameters
input (torch.Tensor) – the input tensor.
dim (int) – the dimension to calculate autocorrelation.
- Returns torch.Tensor
autocorrelation of
input
.
- effective_sample_size(input, chain_dim=0, sample_dim=1)[source]¶
Computes effective sample size of input.
Reference:
- [1] Introduction to Markov Chain Monte Carlo,
Charles J. Geyer
- [2] Stan Reference Manual version 2.18,
Stan Development Team
- Parameters
input (torch.Tensor) – the input tensor.
chain_dim (int) – the chain dimension.
sample_dim (int) – the sample dimension.
- Returns torch.Tensor
effective sample size of
input
.
- resample(input, num_samples, dim=0, replacement=False)[source]¶
Draws
num_samples
samples frominput
at dimensiondim
.- Parameters
input (torch.Tensor) – the input tensor.
num_samples (int) – the number of samples to draw from
input
.dim (int) – dimension to draw from
input
.
- Returns torch.Tensor
samples drawn randomly from
input
.
- quantile(input, probs, dim=0)[source]¶
Computes quantiles of
input
atprobs
. Ifprobs
is a scalar, the output will be squeezed atdim
.- Parameters
input (torch.Tensor) – the input tensor.
probs (list) – quantile positions.
dim (int) – dimension to take quantiles from
input
.
- Returns torch.Tensor
quantiles of
input
atprobs
.
- weighed_quantile(input: torch.Tensor, probs: Union[List[float], Tuple[float, ...], torch.Tensor], log_weights: torch.Tensor, dim: int = 0) torch.Tensor [source]¶
Computes quantiles of weighed
input
samples atprobs
.- Parameters
input (torch.Tensor) – the input tensor.
probs (list) – quantile positions.
log_weights (torch.Tensor) – sample weights tensor.
dim (int) – dimension to take quantiles from
input
.
- Returns torch.Tensor
quantiles of
input
atprobs
.
Example:
>>> from pyro.ops.stats import weighed_quantile >>> import torch >>> input = torch.Tensor([[10, 50, 40], [20, 30, 0]]) >>> probs = torch.Tensor([0.2, 0.8]) >>> log_weights = torch.Tensor([0.4, 0.5, 0.1]).log() >>> result = weighed_quantile(input, probs, log_weights, -1) >>> torch.testing.assert_close(result, torch.Tensor([[40.4, 47.6], [9.0, 26.4]]))
- pi(input, prob, dim=0)[source]¶
Computes percentile interval which assigns equal probability mass to each tail of the interval.
- Parameters
input (torch.Tensor) – the input tensor.
prob (float) – the probability mass of samples within the interval.
dim (int) – dimension to calculate percentile interval from
input
.
- Returns torch.Tensor
quantiles of
input
atprobs
.
- hpdi(input, prob, dim=0)[source]¶
Computes “highest posterior density interval” which is the narrowest interval with probability mass
prob
.- Parameters
input (torch.Tensor) – the input tensor.
prob (float) – the probability mass of samples within the interval.
dim (int) – dimension to calculate percentile interval from
input
.
- Returns torch.Tensor
quantiles of
input
atprobs
.
- waic(input, log_weights=None, pointwise=False, dim=0)[source]¶
Computes “Widely Applicable/Watanabe-Akaike Information Criterion” (WAIC) and its corresponding effective number of parameters.
Reference:
[1] WAIC and cross-validation in Stan, Aki Vehtari, Andrew Gelman
- Parameters
input (torch.Tensor) – the input tensor, which is log likelihood of a model.
log_weights (torch.Tensor) – weights of samples along
dim
.dim (int) – the sample dimension of
input
.
- Returns tuple
tuple of WAIC and effective number of parameters.
- fit_generalized_pareto(X)[source]¶
Given a dataset X assumed to be drawn from the Generalized Pareto Distribution, estimate the distributional parameters k, sigma using a variant of the technique described in reference [1], as described in reference [2].
References [1] ‘A new and efficient estimation method for the generalized Pareto distribution.’ Zhang, J. and Stephens, M.A. (2009). [2] ‘Pareto Smoothed Importance Sampling.’ Aki Vehtari, Andrew Gelman, Jonah Gabry
- Parameters
torch.Tensor – the input data X
- Returns tuple
tuple of floats (k, sigma) corresponding to the fit parameters
- crps_empirical(pred, truth)[source]¶
Computes negative Continuous Ranked Probability Score CRPS* [1] between a set of samples
pred
and true datatruth
. This uses ann log(n)
time algorithm to compute a quantity equal that would naively have complexity quadratic in the number of samplesn
:CRPS* = E|pred - truth| - 1/2 E|pred - pred'| = (pred - truth).abs().mean(0) - (pred - pred.unsqueeze(1)).abs().mean([0, 1]) / 2
Note that for a single sample this reduces to absolute error.
References
- [1] Tilmann Gneiting, Adrian E. Raftery (2007)
Strictly Proper Scoring Rules, Prediction, and Estimation https://www.stat.washington.edu/raftery/Research/PDF/Gneiting2007jasa.pdf
- Parameters
pred (torch.Tensor) – A set of sample predictions batched on rightmost dim. This should have shape
(num_samples,) + truth.shape
.truth (torch.Tensor) – A tensor of true observations.
- Returns
A tensor of shape
truth.shape
.- Return type
- energy_score_empirical(pred: torch.Tensor, truth: torch.Tensor) torch.Tensor [source]¶
Computes negative Energy Score ES* (see equation 22 in [1]) between a set of multivariate samples
pred
and a true data vectortruth
. Running time is quadratic in the number of samplesn
. In case of univariate samples the output coincides with the CRPS:ES* = E|pred - truth| - 1/2 E|pred - pred'|
Note that for a single sample this reduces to the Euclidean norm of the difference between the sample
pred
and thetruth
.This is a strictly proper score so that for
pred
distirbuted according to a distribution \(P\) andtruth
distributed according to a distribution \(Q\) we have \(ES^{*}(P,Q) \ge ES^{*}(Q,Q)\) with equality holding if and only if \(P=Q\), i.e. if \(P\) and \(Q\) have the same multivariate distribution (it is not sufficient for \(P\) and \(Q\) to have the same marginals in order for equality to hold).References
- [1] Tilmann Gneiting, Adrian E. Raftery (2007)
Strictly Proper Scoring Rules, Prediction, and Estimation https://www.stat.washington.edu/raftery/Research/PDF/Gneiting2007jasa.pdf
- Parameters
pred (torch.Tensor) – A set of sample predictions batched on the second leftmost dim. The leftmost dim is that of the multivariate sample.
truth (torch.Tensor) – A tensor of true observations with same shape as
pred
except for the second leftmost dim which can have any value or be omitted.
- Returns
A tensor of shape
truth.shape
.- Return type
Streaming Statistics¶
- class StreamingStats[source]¶
Bases:
abc.ABC
Abstract base class for streamable statistics of trees of tensors.
Derived classes must implelement
update()
,merge()
, andget()
.- abstract update(sample) None [source]¶
Update state from a single sample.
This mutates
self
and returns nothing. Updates should be independent of order, i.e. samples should be exchangeable.- Parameters
sample – A sample value which is a nested dictionary of
torch.Tensor
leaves. This can have arbitrary nesting and shape shape, but assumes shape is constant across calls to.update()
.
- abstract merge(other) pyro.ops.streaming.StreamingStats [source]¶
Select two aggregate statistics, e.g. from different MCMC chains.
This is a pure function: it returns a new
StreamingStats
object and does not modify eitherself
orother
.- Parameters
other – Another streaming stats instance of the same type.
- class StatsOfDict(types: Dict[Hashable, Callable[[], pyro.ops.streaming.StreamingStats]] = {}, default: Callable[[], pyro.ops.streaming.StreamingStats] = <class 'pyro.ops.streaming.CountStats'>)[source]¶
Bases:
pyro.ops.streaming.StreamingStats
Statistics of samples that are dictionaries with constant set of keys.
For example the following are equivalent:
# Version 1. Hand encode statistics. >>> a_stats = CountStats() >>> b_stats = CountMeanStats() >>> a_stats.update(torch.tensor(0.)) >>> b_stats.update(torch.tensor([1., 2.])) >>> summary = {"a": a_stats.get(), "b": b_stats.get()} # Version 2. Collect samples into dictionaries. >>> stats = StatsOfDict({"a": CountStats, "b": CountMeanStats}) >>> stats.update({"a": torch.tensor(0.), "b": torch.tensor([1., 2.])}) >>> summary = stats.get() >>> summary {'a': {'count': 1}, 'b': {'count': 1, 'mean': tensor([1., 2.])}}
- Parameters
default – Default type of statistics of values of the dictionary. Defaults to the inexpensive
CountStats
.types (dict) – Dictionary mapping key to type of statistic that should be recorded for values corresponding to that key.
- class StackStats[source]¶
Bases:
pyro.ops.streaming.StreamingStats
Statistic collecting a stream of tensors into a single stacked tensor.
- update(sample: torch.Tensor) None [source]¶
- class CountStats[source]¶
Bases:
pyro.ops.streaming.StreamingStats
Statistic tracking only the number of samples.
For example:
>>> stats = CountStats() >>> stats.update(torch.randn(3, 3)) >>> stats.get() {'count': 1}
- class CountMeanStats[source]¶
Bases:
pyro.ops.streaming.StreamingStats
Statistic tracking the count and mean of a single
torch.Tensor
.- update(sample: torch.Tensor) None [source]¶
- class CountMeanVarianceStats[source]¶
Bases:
pyro.ops.streaming.StreamingStats
Statistic tracking the count, mean, and (diagonal) variance of a single
torch.Tensor
.- update(sample: torch.Tensor) None [source]¶
State Space Model and GP Utilities¶
- class MaternKernel(nu=1.5, num_gps=1, length_scale_init=None, kernel_scale_init=None)[source]¶
Bases:
pyro.nn.module.PyroModule
Provides the building blocks for representing univariate Gaussian Processes (GPs) with Matern kernels as state space models.
- Parameters
nu (float) – The order of the Matern kernel (one of 0.5, 1.5 or 2.5)
num_gps (int) – the number of GPs
length_scale_init (torch.Tensor) – optional num_gps-dimensional vector of initializers for the length scale
kernel_scale_init (torch.Tensor) – optional num_gps-dimensional vector of initializers for the kernel scale
References
- [1] Kalman Filtering and Smoothing Solutions to Temporal Gaussian Process Regression Models,
Jouni Hartikainen and Simo Sarkka.
- [2] Stochastic Differential Equation Methods for Spatio-Temporal Gaussian Process Regression,
Arno Solin.
- transition_matrix(dt)[source]¶
Compute the (exponentiated) transition matrix of the GP latent space. The resulting matrix has layout (num_gps, old_state, new_state), i.e. this matrix multiplies states from the right.
See section 5 in reference [1] for details.
- Parameters
dt (float) – the time interval over which the GP latent space evolves.
- Returns torch.Tensor
a 3-dimensional tensor of transition matrices of shape (num_gps, state_dim, state_dim).
- stationary_covariance()[source]¶
Compute the stationary state covariance. See Eqn. 3.26 in reference [2].
- Returns torch.Tensor
a 3-dimensional tensor of covariance matrices of shape (num_gps, state_dim, state_dim).
- process_covariance(A)[source]¶
Given a transition matrix A computed with transition_matrix compute the the process covariance as described in Eqn. 3.11 in reference [2].
- Returns torch.Tensor
a batched covariance matrix of shape (num_gps, state_dim, state_dim)
- transition_matrix_and_covariance(dt)[source]¶
Get the transition matrix and process covariance corresponding to a time interval dt.
- Parameters
dt (float) – the time interval over which the GP latent space evolves.
- Returns tuple
(transition_matrix, process_covariance) both 3-dimensional tensors of shape (num_gps, state_dim, state_dim)