fix: relax python 3.14 dependency pins

[stefan-jansen]

Summary:

• drop Python 3.14-specific split pins for pandas, pyarrow, scipy, scikit-learn, statsmodels, and numba
• keep the package source-level minimums unconditional so downstream resolvers can choose newer compatible builds

Verification:

• constrained Python 3.14 resolution with pandas==2.3.3 and scikit-learn==1.6.1 succeeded
• full suite under that constrained environment: 2752 passed, 65 skipped, 24 deselected
• uv build passed

[Copilot]

Pull request overview

Relaxes dependency constraints in pyproject.toml by removing Python 3.14-specific split markers for several scientific/data packages, while documenting the change in the changelog. This keeps the declared floors as unconditional requirements and lets resolvers select the newest compatible builds per platform/Python.

Changes:

• Removed python_version >= '3.14' split dependency entries for pandas, pyarrow, scipy, scikit-learn, statsmodels, and numba in favor of unconditional minimums.
• Added a 0.1.0b8 changelog entry describing the dependency pin relaxation.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
pyproject.toml	Simplifies dependency specifiers by dropping Python 3.14 split markers for core scientific/data dependencies.
CHANGELOG.md	Adds release notes entry for 0.1.0b8 describing the dependency pin change.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.