# Copyright Contributors to the Pyro project.
# SPDX-License-Identifier: Apache-2.0
import torch
from pyro.nn import DenseNN
from .. import constraints
from ..torch_transform import TransformModule
from ..util import copy_docs_from
from .spline import ConditionalSpline, Spline
[docs]@copy_docs_from(TransformModule)
class SplineCoupling(TransformModule):
r"""
An implementation of the coupling layer with rational spline bijections of
linear and quadratic order (Durkan et al., 2019; Dolatabadi et al., 2020).
Rational splines are functions that are comprised of segments that are the ratio
of two polynomials (see :class:`~pyro.distributions.transforms.Spline`).
The spline coupling layer uses the transformation,
:math:`\mathbf{y}_{1:d} = g_\theta(\mathbf{x}_{1:d})`
:math:`\mathbf{y}_{(d+1):D} = h_\phi(\mathbf{x}_{(d+1):D};\mathbf{x}_{1:d})`
where :math:`\mathbf{x}` are the inputs, :math:`\mathbf{y}` are the outputs,
e.g. :math:`\mathbf{x}_{1:d}` represents the first :math:`d` elements of the
inputs, :math:`g_\theta` is either the identity function or an elementwise
rational monotonic spline with parameters :math:`\theta`, and :math:`h_\phi` is
a conditional elementwise spline spline, conditioning on the first :math:`d`
elements.
Example usage:
>>> from pyro.nn import DenseNN
>>> input_dim = 10
>>> split_dim = 6
>>> count_bins = 8
>>> base_dist = dist.Normal(torch.zeros(input_dim), torch.ones(input_dim))
>>> param_dims = [(input_dim - split_dim) * count_bins,
... (input_dim - split_dim) * count_bins,
... (input_dim - split_dim) * (count_bins - 1),
... (input_dim - split_dim) * count_bins]
>>> hypernet = DenseNN(split_dim, [10*input_dim], param_dims)
>>> transform = SplineCoupling(input_dim, split_dim, hypernet)
>>> pyro.module("my_transform", transform) # doctest: +SKIP
>>> flow_dist = dist.TransformedDistribution(base_dist, [transform])
>>> flow_dist.sample() # doctest: +SKIP
:param input_dim: Dimension of the input vector. Despite operating element-wise,
this is required so we know how many parameters to store.
:type input_dim: int
:param split_dim: Zero-indexed dimension :math:`d` upon which to perform input/
output split for transformation.
:param hypernet: a neural network whose forward call returns a tuple of spline
parameters (see :class:`~pyro.distributions.transforms.ConditionalSpline`).
:type hypernet: callable
:param count_bins: The number of segments comprising the spline.
:type count_bins: int
:param bound: The quantity :math:`K` determining the bounding box,
:math:`[-K,K]\times[-K,K]`, of the spline.
:type bound: float
:param order: One of ['linear', 'quadratic'] specifying the order of the spline.
:type order: string
References:
Conor Durkan, Artur Bekasov, Iain Murray, George Papamakarios. Neural
Spline Flows. NeurIPS 2019.
Hadi M. Dolatabadi, Sarah Erfani, Christopher Leckie. Invertible Generative
Modeling using Linear Rational Splines. AISTATS 2020.
"""
domain = constraints.real_vector
codomain = constraints.real_vector
bijective = True
def __init__(
self,
input_dim,
split_dim,
hypernet,
count_bins=8,
bound=3.0,
order="linear",
identity=False,
):
super(SplineCoupling, self).__init__(cache_size=1)
# One part of the input is (optionally) put through an element-wise spline and the other part through a
# conditional one that inputs the first part.
self.lower_spline = Spline(split_dim, count_bins, bound, order)
self.upper_spline = ConditionalSpline(
hypernet, input_dim - split_dim, count_bins, bound, order
)
self.split_dim = split_dim
self.identity = identity
def _call(self, x):
"""
:param x: the input into the bijection
:type x: torch.Tensor
Invokes the bijection x=>y; in the prototypical context of a
:class:`~pyro.distributions.TransformedDistribution` `x` is a sample from
the base distribution (or the output of a previous transform)
"""
x1, x2 = x[..., : self.split_dim], x[..., self.split_dim :]
if not self.identity:
y1 = self.lower_spline(x1)
log_detK = self.lower_spline._cache_log_detJ
else:
y1 = x1
upper_spline = self.upper_spline.condition(x1)
y2 = upper_spline(x2)
log_detJ = upper_spline._cache_log_detJ
if not self.identity:
log_detJ = torch.cat([log_detJ, log_detK], dim=-1)
self._cache_log_detJ = log_detJ
return torch.cat([y1, y2], dim=-1)
def _inverse(self, y):
"""
:param y: the output of the bijection
:type y: torch.Tensor
Inverts y => x. Uses a previously cached inverse if available,
otherwise performs the inversion afresh.
"""
y1, y2 = y[..., : self.split_dim], y[..., self.split_dim :]
if not self.identity:
x1 = self.lower_spline._inv_call(y1)
log_detK = self.lower_spline._cache_log_detJ
else:
x1 = y1
upper_spline = self.upper_spline.condition(x1)
x2 = upper_spline._inv_call(y2)
log_detJ = upper_spline._cache_log_detJ
if not self.identity:
log_detJ = torch.cat([log_detJ, log_detK], dim=-1)
self._cache_log_detJ = log_detJ
return torch.cat([x1, x2], dim=-1)
[docs] def log_abs_det_jacobian(self, x, y):
"""
Calculates the elementwise determinant of the log jacobian
"""
x_old, y_old = self._cached_x_y
if x is not x_old or y is not y_old:
# This call to the parent class Transform will update the cache
# as well as calling self._call and recalculating y and log_detJ
self(x)
return self._cache_log_detJ.sum(-1)
[docs]def spline_coupling(
input_dim, split_dim=None, hidden_dims=None, count_bins=8, bound=3.0
):
"""
A helper function to create a
:class:`~pyro.distributions.transforms.SplineCoupling` object for consistency
with other helpers.
:param input_dim: Dimension of input variable
:type input_dim: int
"""
if split_dim is None:
split_dim = input_dim // 2
if hidden_dims is None:
hidden_dims = [input_dim * 10, input_dim * 10]
nn = DenseNN(
split_dim,
hidden_dims,
param_dims=[
(input_dim - split_dim) * count_bins,
(input_dim - split_dim) * count_bins,
(input_dim - split_dim) * (count_bins - 1),
(input_dim - split_dim) * count_bins,
],
)
return SplineCoupling(input_dim, split_dim, nn, count_bins, bound)