Fixes#144196
Extends #144106 and #144110
## Open Problems:
- [ ] Annotating with `numbers.Number` is a bad idea, should consider using `float`, `SupportsFloat` or some `Procotol`. https://github.com/pytorch/pytorch/pull/144197#discussion_r1903324769
# Notes
- `beta.py`: needed to add `type: ignore` since `broadcast_all` is untyped.
- `categorical.py`: converted `else` branches of mutually exclusive arguments to `if` branch[^2].
- ~~`dirichlet.py`: replaced `axis` with `dim` arguments.~~ #144402
- `gemoetric.py`: converted `else` branches of mutually exclusive arguments to `if` branch[^2].
- ~~`independent.py`: fixed bug in `Independent.__init__` where `tuple[int, ...]` could be passed to `Distribution.__init__` instead of `torch.Size`.~~ **EDIT:** turns out the bug is related to typing of `torch.Size`. #144218
- `independent.py`: made `Independent` a generic class of its base distribution.
- `multivariate_normal.py`: converted `else` branches of mutually exclusive arguments to `if` branch[^2].
- `relaxed_bernoulli.py`: added class-level type hint for `base_dist`.
- `relaxed_categorical.py`: added class-level type hint for `base_dist`.
- ~~`transforms.py`: Added missing argument to docstring of `ReshapeTransform`~~ #144401
- ~~`transforms.py`: Fixed bug in `AffineTransform.sign` (could return `Tensor` instead of `int`).~~ #144400
- `transforms.py`: Added `type: ignore` comments to `AffineTransform.log_abs_det_jacobian`[^1]; replaced `torch.abs(scale)` with `scale.abs()`.
- `transforms.py`: Added `type: ignore` comments to `AffineTransform.__eq__`[^1].
- `transforms.py`: Fixed type hint on `CumulativeDistributionTransform.domain`. Note that this is still an LSP violation, because `Transform.domain` is defined as `Constraint`, but `Distribution.domain` is defined as `Optional[Constraint]`.
- skipped: `constraints.py`, `constraints_registry.py`, `kl.py`, `utils.py`, `exp_family.py`, `__init__.py`.
## Remark
`TransformedDistribution`: `__init__` uses the check `if reinterpreted_batch_ndims > 0:`, which can lead to the creation of `Independent` distributions with only 1 component. This results in awkward code like `base_dist.base_dist` in `LogisticNormal`.
```python
import torch
from torch.distributions import *
b1 = Normal(torch.tensor([0.0]), torch.tensor([1.0]))
b2 = MultivariateNormal(torch.tensor([0.0]), torch.eye(1))
t = StickBreakingTransform()
d1 = TransformedDistribution(b1, t)
d2 = TransformedDistribution(b2, t)
print(d1.base_dist) # Independent with 1 dimension
print(d2.base_dist) # MultivariateNormal
```
One could consider changing this to `if reinterpreted_batch_ndims > 1:`.
[^1]: Usage of `isinstance(value, numbers.Real)` leads to problems with static typing, as the `numbers` module is not supported by `mypy` (see <https://github.com/python/mypy/issues/3186>). This results in us having to add type-ignore comments in several places
[^2]: Otherwise, we would have to add a bunch of `type: ignore` comments to make `mypy` happy, as it isn't able to perform the type narrowing. Ideally, such code should be replaced with structural pattern matching once support for Python 3.9 is dropped.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/144197
Approved by: https://github.com/malfet
Co-authored-by: Aaron Gokaslan <aaronGokaslan@gmail.com>
Fixes#76772, #144196
Extends #144106
- added type annotations to `lazy_property`.
- added type annotation to all `@property` and `@lazy_property` inside `torch.distributions` module.
- added simply type-check unit test to ensure type inference is working.
- replaced deprecated annotations like `typing.List` with the corresponding counterpart.
- simplified `torch.Tensor` hints with plain `Tensor`, otherwise signatures can become very verbose.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/144110
Approved by: https://github.com/Skylion007
This is a suggestion for a minor modification.
The line `log_normalization[self.total_count + value == 0.] = 0.` prevents Jit compilation when the condition occurs, with the error message
`RuntimeError: a view of a leaf Variable that requires grad is being used in an in-place operation.`
I propose an alternative that does not involve in-place operations. It uses the function `nan_to_num()` to replace infinite values by 0 where `self.total_count + value == 0.` while leaving `nan` and `-inf` as they are. Readability is suboptimal because the code does not replace nan with numbers, but I could not find a function that only replaces infinite values.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/96748
Approved by: https://github.com/fritzo, https://github.com/soulitzer
This PR fixes#69466 and introduces some other minor changes. Tests are somewhat more involved because a reference implementation in `scipy` is not available; tests proceed differently for discrete and continuous distributions.
For continuous distributions, we evaluate the gradient of the `log_prob` at the mode. Tests pass if the gradient is zero OR (the mode is at the boundary of the support of the distribution AND the `log_prob` decreases as we move away from the boundary to the interior of the support).
For discrete distributions, the notion of a gradient is not well defined. We thus "look" ahead and behind one step (e.g. if the mode of a Poisson distribution is 9, we consider 8 and 10). If the step ahead/behind is still within the support of the distribution, we assert that the `log_prob` is smaller than at the mode.
For one-hot encoded distributions (currently just `OneHotCategorical`), we evaluate the underlying mode (i.e. encoded as an integral tensor), "advance" by one label to get another sample that should have lower probability using `other = (mode + 1) % event_size` and re-encode as one-hot. The resultant `other` sample should have lower probability than the mode.
Furthermore, Gamma, half Cauchy, and half normal distributions have their support changed from positive to nonnegative. This change is necessary because the mode of the "half" distributions is zero, and the mode of the gamma distribution is zero for `concentration <= 1`.
cc @fritzo
Pull Request resolved: https://github.com/pytorch/pytorch/pull/76690
Approved by: https://github.com/neerajprad
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/63192
**Summary**
This reverts commit 402caaeba513929dcfe12df183c764b0ef43f688. As per the
dicussion in #62178, this commit was not needed.
**Test Plan**
Continuous integration.
Test Plan: Imported from OSS
Reviewed By: VitalyFedyunin
Differential Revision: D30293202
Pulled By: SplitInfinity
fbshipit-source-id: 91ee7ad0523a9880605d83fe9712c39df67384a8
Summary:
Fixes https://github.com/pytorch/pytorch/issues/47123
Follows https://github.com/pyro-ppl/pyro/pull/2701
This turns on `Distribution` validation by default. The motivation is to favor beginners by providing helpful error messages. Advanced users focused on speed can disable validation by calling
```py
torch.distributions.Distribution.set_default_validate_args(False)
```
or by disabling individual distribution validation via `MyDistribution(..., validate_args=False)`.
In practice I have found many beginners forget or do not know about validation. Therefore I have [enabled it by default](https://github.com/pyro-ppl/pyro/pull/2701) in Pyro. I believe PyTorch could also benefit from this change. Indeed validation caught a number of bugs in `.icdf()` methods, in tests, and in PPL benchmarks, all of which have been fixed in this PR.
## Release concerns
- This may slightly slow down some models. Concerned users may disable validation.
- This may cause new `ValueErrors` in models that rely on unsupported behavior, e.g. `Categorical.log_prob()` applied to continuous-valued tensors (only {0,1}-valued tensors are supported).
We should clearly note this change in release notes.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/48743
Reviewed By: heitorschueroff
Differential Revision: D25304247
Pulled By: neerajprad
fbshipit-source-id: 8d50f28441321ae691f848c55f71aa80cb356b41
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/45693
**Summary**
This commit updates the docstring for
`torch.distributions.NegativeBinomial` to better match actual behaviour.
In particular, the parameter currently documented as probability of
success is actually probability of failure.
**Test Plan**
1) Ran the code from the issue to make sure this is still an issue (it
is)
2) `make html` and viewed the docs in a browser.
*Before*
<img width="879" alt="Captura de Pantalla 2020-10-01 a la(s) 1 35 28 p m" src="https://user-images.githubusercontent.com/4392003/94864456-db3a5680-03f0-11eb-977e-3bab0fb9c206.png">
*After*
<img width="877" alt="Captura de Pantalla 2020-10-01 a la(s) 2 12 24 p m" src="https://user-images.githubusercontent.com/4392003/94864478-e42b2800-03f0-11eb-965a-51493ca27c80.png">
**Fixes**
This commit closes#42449.
Test Plan: Imported from OSS
Reviewed By: robieta
Differential Revision: D24071048
Pulled By: SplitInfinity
fbshipit-source-id: d345b4de721475dbe26233e368af62eb57a47970
Summary:
I have experienced that sometimes both were in `__dict__`, but it chose to copy `probs` which loses precision over `logits`. This is especially important when training (bayesian) neural networks or doing other type of optimization, since the loss is heavily affected.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/18614
Differential Revision: D14793486
Pulled By: ezyang
fbshipit-source-id: d4ff5e34fbb4021ea9de9f58af09a7de00d80a63
Summary:
This adds a `.expand` method for distributions that is akin to the `torch.Tensor.expand` method for tensors. It returns a new distribution instance with batch dimensions expanded to the desired `batch_shape`. Since this calls `torch.Tensor.expand` on the distribution's parameters, it does not allocate new memory for the expanded distribution instance's parameters.
e.g.
```python
>>> d = dist.Normal(torch.zeros(100, 1), torch.ones(100, 1))
>>> d.sample().shape
torch.Size([100, 1])
>>> d.expand([100, 10]).sample().shape
torch.Size([100, 10])
```
We have already been using the `.expand` method in Pyro in our [patch](https://github.com/uber/pyro/blob/dev/pyro/distributions/torch.py#L10) of `torch.distributions`. We use this in our models to enable dynamic broadcasting. This has also been requested by a few users on the distributions slack, and we believe will be useful to the larger community.
Note that currently, there is no convenient and efficient way to expand distribution instances:
- Many distributions use `TransformedDistribution` (or wrap over another distribution instance. e.g. `OneHotCategorical` uses a `Categorical` instance) under the hood, or have lazy parameters. This makes it difficult to collect all the relevant parameters, broadcast them and construct new instances.
- In the few cases where this is even possible, the resulting implementation would be inefficient since we will go through a lot of broadcasting and args validation logic in `__init__.py` that can be avoided.
The `.expand` method allows for a safe and efficient way to expand distribution instances. Additionally, this bypasses `__init__.py` (using `__new__` and populating relevant attributes) since we do not need to do any broadcasting or args validation (which was already done when the instance was first created). This can result in significant savings as compared to constructing new instances via `__init__` (that said, the `sample` and `log_prob` methods will probably be the rate determining steps in many applications).
e.g.
```python
>>> a = dist.Bernoulli(torch.ones([10000, 1]), validate_args=True)
>>> %timeit a.expand([10000, 100])
15.2 µs ± 224 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
>>> %timeit dist.Bernoulli(torch.ones([10000, 100]), validate_args=True)
11.8 ms ± 153 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
```
cc. fritzo, apaszke, vishwakftw, alicanb
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11341
Differential Revision: D9728485
Pulled By: soumith
fbshipit-source-id: 3b94c23bc6a43ee704389e6287aa83d1e278d52f
Summary:
`__repr__` currently fails for distributions with lazy attributes in PyTorch master, throwing a `KeyError`. This fixes the issue.
**Additionally:**
- Added `logits` to `arg_constraints` for distributions that accept either `probs` or `logits`. This is both to have `__repr__` display the `logits` param when available, and to be able to do validation checks (e.g. NaN checks) when the logit parametrization is used. fritzo, alicanb - I think there were reasons why we had not done so in the first place, but I am unable to recall now. It passes all the tests, but let me know if there is something that I am missing at the moment.
- There are certain distributions, e.g. `OneHotCategorical` which won't show any parameters because it uses a `categorical` instance under the hood and neither `logits` / `probs` in `arg_constraints` are present in the instance's `__dict__`. This isn't addressed in this PR.
cc. vishwakftw, fritzo, nadavbh12, apaszke
Pull Request resolved: https://github.com/pytorch/pytorch/pull/11263
Differential Revision: D9654959
Pulled By: apaszke
fbshipit-source-id: 16f5b20243fe8e2c13e9c528050d4df0b8ea6e45
