Reject complex object sigma-point inputs

[FlorianPfaff]

Summary

• detect complex entries hidden inside NumPy object-dtype arrays before backend conversion
• preserve the existing ValueError contract for complex sigma-point means and covariances
• retain support for real-valued object arrays
• add focused regressions for both Julier and Merwe sigma-point schemes

Root cause

_has_complex_dtype() only inspected the array dtype. A normal complex array has dtype kind c, but an object array such as np.array([1 + 1j, 0], dtype=object) has dtype kind O. The guard therefore returned false, and the subsequent conversion to float64 leaked a backend TypeError instead of the documented input-validation ValueError.

Fix

For object-dtype inputs, inspect the flattened entries for Python or NumPy complex scalar types. Native complex dtypes and backend dtypes whose names contain complex keep their existing fast path; non-object real dtypes remain unchanged.

Impact

Sigma-point callers now receive consistent, parameter-specific validation errors regardless of whether complex values arrive in native complex arrays or object arrays. Valid real object arrays continue to work.

Validation

• reproduced the pre-fix TypeError for object-dtype complex means and covariances
• verified the patched path raises ValueError with x must contain real values or P must contain real values
• verified both JulierSigmaPoints and MerweScaledSigmaPoints
• verified real-valued object arrays still generate sigma points
• python -m py_compile passes for the modified module
• branch comparison: two commits ahead, zero behind; only the sampler and focused regression file changed

The full backend matrix is delegated to GitHub Actions.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ COPYPASTE	jscpd	yes		no	no	22.65s
✅ JSON	prettier	7	0	0	0	1.05s
✅ JSON	v8r	7		0	0	2.95s
✅ MARKDOWN	markdownlint	68	0	0	0	1.95s
✅ MARKDOWN	markdown-table-formatter	68	0	0	0	0.63s
✅ PYTHON	black	1490	182	0	0	86.41s
✅ PYTHON	isort	1490	328	0	0	2.51s
✅ REPOSITORY	betterleaks	yes		no	no	2.74s
✅ REPOSITORY	checkov	yes		no	no	53.12s
✅ REPOSITORY	gitleaks	yes		no	no	13.61s
✅ REPOSITORY	git_diff	yes		no	no	0.28s
✅ REPOSITORY	secretlint	yes		no	no	59.24s
✅ REPOSITORY	syft	yes		no	no	6.56s
✅ REPOSITORY	trivy-sbom	yes		no	no	5.54s
✅ REPOSITORY	trufflehog	yes		no	no	30.08s
✅ YAML	prettier	11	0	0	0	0.75s
✅ YAML	v8r	11		0	0	10.15s
✅ YAML	yamllint	11		0	0	0.64s

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