Skip to content

Fix PyTorch multivariate_normal array-like inputs#2228

Open
FlorianPfaff wants to merge 3 commits into
mainfrom
codex/fix-pytorch-mvn-list-inputs
Open

Fix PyTorch multivariate_normal array-like inputs#2228
FlorianPfaff wants to merge 3 commits into
mainfrom
codex/fix-pytorch-mvn-list-inputs

Conversation

@FlorianPfaff
Copy link
Copy Markdown
Owner

Summary

  • coerce PyTorch random.multivariate_normal mean/covariance arguments to tensors before constructing torch.distributions.MultivariateNormal
  • preserve tensor device placement and floating precision when either argument is already a tensor
  • add backend-random regression coverage for NumPy-style Python sequence inputs

Rationale

The NumPy and JAX random backends accept ordinary array-like inputs for random.multivariate_normal, e.g. random.multivariate_normal([0.0, 0.0], [[1.0, 0.0], [0.0, 1.0]], size=(6,)). The PyTorch backend passed these lists directly into torch.distributions.MultivariateNormal, which fails because PyTorch expects tensor arguments.

Validation

  • Connector compare confirms this branch is two commits ahead of current main and behind by zero.
  • Changes are limited to src/pyrecest/_backend/pytorch/random.py and tests/test_backend_random.py.
  • Local smoke test of the new PyTorch conversion helper with list inputs and tensor dtype preservation passed in the sandbox's installed PyTorch.
  • Full repository tests were not run because the sandbox cannot clone/install the repository from GitHub; PR CI should run the backend matrix.

@FlorianPfaff FlorianPfaff enabled auto-merge (squash) May 24, 2026 07:31
@github-actions
Copy link
Copy Markdown
Contributor

github-actions Bot commented May 24, 2026

⚠️MegaLinter analysis: Success with warnings

Descriptor Linter Files Fixed Errors Warnings Elapsed time
✅ COPYPASTE jscpd yes no no 35.15s
✅ JSON prettier 7 0 0 0 0.92s
✅ JSON v8r 7 0 0 4.98s
⚠️ MARKDOWN markdownlint 66 0 1 0 1.8s
✅ MARKDOWN markdown-table-formatter 66 0 0 0 0.79s
✅ PYTHON black 758 3 0 0 36.94s
✅ PYTHON isort 758 3 0 0 2.17s
✅ REPOSITORY checkov yes no no 34.76s
✅ REPOSITORY gitleaks yes no no 14.52s
✅ REPOSITORY git_diff yes no no 0.09s
✅ REPOSITORY secretlint yes no no 24.09s
✅ REPOSITORY syft yes no no 6.62s
✅ REPOSITORY trivy-sbom yes no no 5.47s
✅ REPOSITORY trufflehog yes no no 19.89s
✅ YAML prettier 10 0 0 0 0.65s
✅ YAML v8r 10 0 0 10.43s
✅ YAML yamllint 10 0 0 0.44s

Detailed Issues

⚠️ MARKDOWN / markdownlint - 1 error
.github/pull_request_template.md:1 error MD041/first-line-heading/first-line-h1 First line in a file should be a top-level heading [Context: "## Summary"]

Notices

📣 MegaLinter 9.5.0 is out! Discover the new features and security recommendations in the release announcement. (Skip this info by defining SECURITY_SUGGESTIONS: false)

See detailed reports in MegaLinter artifacts

Your project could benefit from a custom flavor, which would allow you to run only the linters you need, and thus improve runtime performances. (Skip this info by defining FLAVOR_SUGGESTIONS: false)

  • Documentation: Custom Flavors
  • Command: npx mega-linter-runner@9.5.0 --custom-flavor-setup --custom-flavor-linters PYTHON_BLACK,PYTHON_ISORT,COPYPASTE_JSCPD,JSON_V8R,JSON_PRETTIER,MARKDOWN_MARKDOWNLINT,MARKDOWN_MARKDOWN_TABLE_FORMATTER,REPOSITORY_CHECKOV,REPOSITORY_GIT_DIFF,REPOSITORY_GITLEAKS,REPOSITORY_SECRETLINT,REPOSITORY_SYFT,REPOSITORY_TRIVY_SBOM,REPOSITORY_TRUFFLEHOG,YAML_PRETTIER,YAML_YAMLLINT,YAML_V8R

MegaLinter is graciously provided by OX Security
Show us your support by starring ⭐ the repository

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant