Fix PyTorch reduction keepdims contract

[FlorianPfaff]

Summary

• Wrap PyTorch public reduction exports (any, all, max/amax, min/amin, prod) so they accept NumPy-style keepdims.
• Preserve tuple/negative-axis validation and reshape reduced outputs back to singleton reduced axes when keepdims=True.
• Add portable regression coverage, including a forced PYRECEST_BACKEND=pytorch subprocess check.

Rationale

Portable PyRecEst code can call NumPy/JAX reductions with keepdims=True, but the PyTorch backend facade rejected that keyword for several reductions. This caused backend-specific failures for otherwise valid reduction calls.

Validation

• Connector compare confirmed the branch is 2 commits ahead of main and 0 behind before opening this PR.
• Added tests/test_backend_reduction_keepdims_contract.py for default-backend and forced-PyTorch coverage.
• Full local pytest was not run because this execution sandbox cannot resolve github.com; PR CI should run the full matrix.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ COPYPASTE	jscpd	yes		no	no	50.69s
✅ JSON	prettier	7	0	0	0	1.14s
✅ JSON	v8r	7		0	0	4.35s
✅ MARKDOWN	markdownlint	66	0	0	0	2.18s
✅ MARKDOWN	markdown-table-formatter	66	0	0	0	0.75s
✅ PYTHON	black	764	2	0	0	40.86s
✅ PYTHON	isort	764	2	0	0	3.49s
✅ REPOSITORY	checkov	yes		no	no	47.57s
✅ REPOSITORY	gitleaks	yes		no	no	16.36s
✅ REPOSITORY	git_diff	yes		no	no	0.13s
✅ REPOSITORY	secretlint	yes		no	no	22.13s
✅ REPOSITORY	syft	yes		no	no	4.25s
✅ REPOSITORY	trivy-sbom	yes		no	no	7.51s
✅ REPOSITORY	trufflehog	yes		no	no	24.26s
✅ YAML	prettier	11	0	0	0	0.68s
✅ YAML	v8r	11		0	0	11.78s
✅ YAML	yamllint	11		0	0	0.55s

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

Show us your support by starring ⭐ the repository