stats precomp

[leoglonz]

Issue Addressed

Fixes a runtime AttributeError in _normalize and adds norm stats precomputation to reduce per-step overhead in the MTS BMI.

Description

• Precompute normalization statistics at init (_precompute_norm_stats): converts all norm_stats mean/std lists from JSON into numpy arrays once during initialize(), rather than on every call to _normalize.
• Robust _normalize: wraps stat lookups in np.asarray() so the method handles both pre-converted arrays and raw lists (e.g., when norm_stats is set directly in tests), fixing an AttributeError: 'list' object has no attribute 'ndim'.

Type of Change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Code cleanup/refactor
• Documentation update

Other (please specify):

Checklist

• Branch is up to date with master
• Updated tests or added new tests
• Tests & pre-commit hooks pass
• Updated documentation (if applicable)
• Code follows established style and conventions