# Copyright (c) 2017-2019 Uber Technologies, Inc.
# SPDX-License-Identifier: Apache-2.0
from typing import TYPE_CHECKING, List, Optional
from pyro.distributions.torch_distribution import TorchDistributionMixin
from pyro.poutine.messenger import Messenger
from pyro.util import ignore_jit_warnings
if TYPE_CHECKING:
from pyro.poutine.runtime import Message
[docs]class BroadcastMessenger(Messenger):
"""
Automatically broadcasts the batch shape of the stochastic function
at a sample site when inside a single or nested plate context.
The existing `batch_shape` must be broadcastable with the size
of the :class:`~pyro.plate` contexts installed in the
`cond_indep_stack`.
Notice how `model_automatic_broadcast` below automates expanding of
distribution batch shapes. This makes it easy to modularize a
Pyro model as the sub-components are agnostic of the wrapping
:class:`~pyro.plate` contexts.
>>> def model_broadcast_by_hand():
... with IndepMessenger("batch", 100, dim=-2):
... with IndepMessenger("components", 3, dim=-1):
... sample = pyro.sample("sample", dist.Bernoulli(torch.ones(3) * 0.5)
... .expand_by(100))
... assert sample.shape == torch.Size((100, 3))
... return sample
>>> @poutine.broadcast
... def model_automatic_broadcast():
... with IndepMessenger("batch", 100, dim=-2):
... with IndepMessenger("components", 3, dim=-1):
... sample = pyro.sample("sample", dist.Bernoulli(torch.tensor(0.5)))
... assert sample.shape == torch.Size((100, 3))
... return sample
"""
@staticmethod
@ignore_jit_warnings(["Converting a tensor to a Python boolean"])
def _pyro_sample(msg: "Message") -> None:
"""
:param msg: current message at a trace site.
"""
if (
msg["done"]
or msg["type"] != "sample"
or not isinstance(msg["fn"], TorchDistributionMixin)
):
return
dist = msg["fn"]
actual_batch_shape = dist.batch_shape
target_batch_shape = [
None if size == 1 else size for size in actual_batch_shape
]
for f in msg["cond_indep_stack"]:
if f.dim is None or f.size == -1:
continue
assert f.dim < 0
prefix_batch_shape: List[Optional[int]] = [None] * (
-f.dim - len(target_batch_shape)
)
target_batch_shape = prefix_batch_shape + target_batch_shape
if (
target_batch_shape[f.dim] is not None
and target_batch_shape[f.dim] != f.size
):
raise ValueError(
"Shape mismatch inside plate('{}') at site {} dim {}, {} vs {}".format(
f.name,
msg["name"],
f.dim,
f.size,
target_batch_shape[f.dim],
)
)
target_batch_shape[f.dim] = f.size
# Starting from the right, if expected size is None at an index,
# set it to the actual size if it exists, else 1.
for i in range(-len(target_batch_shape) + 1, 1):
if target_batch_shape[i] is None:
target_batch_shape[i] = (
actual_batch_shape[i] if len(actual_batch_shape) >= -i else 1
)
msg["fn"] = dist.expand(target_batch_shape)
if msg["fn"].has_rsample != dist.has_rsample:
msg["fn"].has_rsample = dist.has_rsample # copy custom attribute