Preserve Complex Bingham single-point autograd

[FlorianPfaff]

Bug

ComplexBinghamDistribution.pdf(...) keeps backend-native values for batched evaluation, but the single-point path returned float(p[0]).

On the PyTorch backend, a complex query tensor with requires_grad=True therefore produced a Python float. That conversion silently detached the likelihood from the computation graph, so callers could not differentiate a single-point Complex Bingham density even though the equivalent batched path remained differentiable.

Fix

Return the backend scalar p[0] directly for single-point evaluation instead of converting it to Python float.

This preserves the numerical value while retaining the active backend's scalar type and PyTorch autograd history.

Regression coverage

Add a PyTorch-backend regression that evaluates a unit-norm complex point with requires_grad=True and verifies:

• the result is a scalar PyTorch tensor;
• the result is float64 on the same device as the query;
• the density is finite and positive;
• the density remains differentiable;
• backpropagation reaches the complex query with finite, nonzero gradients.

Validation

• independently reproduced that float(...) turns a differentiable PyTorch scalar into a plain Python float and prevents .backward();
• verified the backend scalar path backpropagates finite, nonzero gradients with PyTorch 2.10;
• final diff is limited to a one-line production fix and one focused 43-line regression module;
• branch is based directly on current main (f196ec41d1d8681e279936db9f1a0a3abfa276e7) and is 2 commits ahead / 0 behind;
• GitHub Actions is authoritative for the full backend, lint, packaging, documentation, and integration matrix.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ COPYPASTE	jscpd	yes		no	no	29.16s
✅ JSON	prettier	7	0	0	0	1.32s
✅ JSON	v8r	7		0	0	4.5s
✅ MARKDOWN	markdownlint	68	0	0	0	1.8s
✅ MARKDOWN	markdown-table-formatter	68	0	0	0	0.89s
✅ PYTHON	black	1867	87	0	0	97.52s
✅ PYTHON	isort	1867	141	0	0	2.81s
✅ REPOSITORY	betterleaks	yes		no	no	2.61s
✅ REPOSITORY	checkov	yes		no	no	54.16s
✅ REPOSITORY	gitleaks	yes		no	no	17.33s
✅ REPOSITORY	git_diff	yes		no	no	0.62s
✅ REPOSITORY	secretlint	yes		no	no	83.49s
✅ REPOSITORY	syft	yes		no	no	6.16s
✅ REPOSITORY	trivy-sbom	yes		no	no	5.88s
✅ REPOSITORY	trufflehog	yes		no	no	32.23s
✅ YAML	prettier	11	0	0	0	0.83s
✅ YAML	v8r	11		0	0	10.81s
✅ YAML	yamllint	11		0	0	0.6s

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@9.6.0 --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_BETTERLEAKS,REPOSITORY_SECRETLINT,REPOSITORY_SYFT,REPOSITORY_TRIVY_SBOM,REPOSITORY_TRUFFLEHOG,YAML_PRETTIER,YAML_YAMLLINT,YAML_V8R

Show us your support by starring ⭐ the repository