Constraints symmetric #68240
Closed
Constraints symmetric #68240
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
CI Flow Status⚛️ CI FlowRuleset - Version:
You can add a comment to the PR and tag @pytorchbot with the following commands: # ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun
# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slowFor more information, please take a look at the CI Flow Wiki. |
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit dfdfa3a (more details on the Dr. CI page):
🕵️ 19 new failures recognized by patternsThe following CI failures do not appear to be due to upstream breakages:
|
saketh-are
added
the
triaged
|
Hello, @albanD. May I have a chance to be reviewed my PR? I am trying to contribute to PyTorch by implementing distributions for symmetric matrices like Wishart distribution and Inverse Wishart distribution. Although there is a LKJ distribution for the cholesky decomposition of correlation matrices, it only represents equivalence to restricted form of Wishart distribution. Thus, I started implementing Wishart distribution and Inverse Wishart distribution seperately. Before sending the PR of distributions, I added a short code about the 'torch.distributions.constraints.symmetric', which was not included in 'torch.distributions.constraints' previously. i.e., 'torch.distributions.constraints' contains module like 'positive_definite' constraints, but it just assumes symmetricity of the input matrix. (Link) So, I think it will be better if we have symmetric constraint in PyTorch. We may further utilize it like Thanks. |
|
Sure, @fritzo would you have some time to take a look? |
…ch into constraints_symmetric
nonconvexopt
requested review from
BowenBao,
H-Huang,
IvanYashchuk,
bowangbj,
ezyang,
janeyx99,
jbschlosser,
jeffdaily,
jithunnair-amd,
lezcano,
mingzhe09088 and
mrshenli
as code owners
nonconvexopt
requested review from
nikitaved,
pritamdamania87,
rohan-varma,
seemethere,
shubhambhokare1,
soulitzer,
zhaojuanmao and
zhouzhuojie
as code owners
|
Sorry for triggering review request. I selected wrong commit to revert. It is my bad. |
Fix revert mistake
3 tasks
|
Sorry for the confusion. Making another PR. |
facebook-github-bot
pushed a commit
that referenced
this pull request
Nov 24, 2021
…he tensor is symmetric at last 2 dimension. (#68644) Summary: Implemented submodule for #68050 Opened cleaned, final version of PR for #68240 Explanation: I am trying to contribute to PyTorch by implementing distributions for symmetric matrices like Wishart distribution and Inverse Wishart distribution. Although there is a LKJ distribution for the Cholesky decomposition of correlation matrices, it only represents equivalence to restricted form of Wishart distribution. [https://arxiv.org/abs/1809.04746](https://arxiv.org/abs/1809.04746) Thus, I started implementing Wishart distribution and Inverse Wishart distribution seperately. I added a short code about the 'torch.distributions.constraints.symmetric', which was not included in 'torch.distributions.constraints' previously. i.e., 'torch.distributions.constraints' contains module like 'positive_definite' constraints, but it just assumes symmetricity of the input matrix. [Link](https://github.com/pytorch/pytorch/blob/1adeeabdc0c8832420c091c5c668843768530d7f/torch/distributions/constraints.py#L466) So, I think it will be better if we have constraint checking symmetricity of the tensors in PyTorch. We may further utilize it like `constraints.stack([constraints.symmetric, constraints.positive_definite])` for the constraint of the covariance matrix in Multivariate Normal distribution, for example, to check if the random matrix is a symmetric positive definite matrix. cc fritzo neerajprad alicanb nikitaved Pull Request resolved: #68644 Reviewed By: jbschlosser Differential Revision: D32599540 Pulled By: neerajprad fbshipit-source-id: 9227f7e9931834a548a88da69e4f2e9af7732cfe
PaliC
added a commit
that referenced
this pull request
Nov 30, 2021
…he tensor is symmetric at last 2 dimension. (#68644) Summary: Implemented submodule for #68050 Opened cleaned, final version of PR for #68240 Explanation: I am trying to contribute to PyTorch by implementing distributions for symmetric matrices like Wishart distribution and Inverse Wishart distribution. Although there is a LKJ distribution for the Cholesky decomposition of correlation matrices, it only represents equivalence to restricted form of Wishart distribution. [https://arxiv.org/abs/1809.04746](https://arxiv.org/abs/1809.04746) Thus, I started implementing Wishart distribution and Inverse Wishart distribution seperately. I added a short code about the 'torch.distributions.constraints.symmetric', which was not included in 'torch.distributions.constraints' previously. i.e., 'torch.distributions.constraints' contains module like 'positive_definite' constraints, but it just assumes symmetricity of the input matrix. [Link](https://github.com/pytorch/pytorch/blob/1adeeabdc0c8832420c091c5c668843768530d7f/torch/distributions/constraints.py#L466) So, I think it will be better if we have constraint checking symmetricity of the tensors in PyTorch. We may further utilize it like `constraints.stack([constraints.symmetric, constraints.positive_definite])` for the constraint of the covariance matrix in Multivariate Normal distribution, for example, to check if the random matrix is a symmetric positive definite matrix. cc fritzo neerajprad alicanb nikitaved Reviewed By: jbschlosser Differential Revision: D32599540 Pulled By: neerajprad fbshipit-source-id: 9227f7e9931834a548a88da69e4f2e9af7732cfe [ghstack-poisoned]
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Opened PR again.
Implemented Symmetric matrices 'torch.distributions.constraints' for implementing