Fix numpy and pandas compatibility

[martin-springer]

This pull request focuses on improving compatibility with recent versions of pandas (3.0+), numpy (2.x), and xgboost (3.x), while also removing deprecated normalization functions and addressing several minor bugs and deprecations. The main changes involve updating internal calculations to be resolution-agnostic, cleaning up deprecated code, and ensuring the codebase works with the latest dependencies.

Dependency and compatibility updates:

• Removed upper version restrictions for pandas and numpy in setup.py to support pandas 3.0 and numpy 2.x; updated related requirements and changelog documentation.
• Fixed pandas 3.0 compatibility throughout normalization.py by replacing .view('int64') with .total_seconds() for datetime calculations, making code resolution-agnostic and robust to changes in pandas datetime resolution. [1] [2] [3] [4] [5] [6]
• Fixed numpy 2.x compatibility in soiling.py by using .item() and explicit indexing for scalar extraction, and replaced deprecated implicit conversions. [1] [2]

Deprecations and removals:

• Removed deprecated normalization.delta_index and normalization.check_series_frequency functions (deprecated since v2.0.0); their private helpers remain for internal use. [1] [2] [3] [4]

Bug fixes and minor enhancements:

• Ensured datetime resolution preservation in normalization.interpolate() so output matches input resolution (e.g., microseconds vs nanoseconds).
• Fixed xgboost 3.x compatibility in filtering.xgboost_clip_filter() by using xgb.DMatrix with explicit feature names for model prediction. [1] [2]
• Fixed pandas 4.0 deprecation warnings by switching Timedelta strings from lowercase 'd' to uppercase 'D' and using axis= keyword for DataFrame aggregation methods. [1] [2] [3] [4] [5]

Issues
Closes #472
Closes #445

These changes collectively ensure that the codebase remains compatible with the latest versions of key dependencies and removes legacy code, while also addressing minor bugs and improving code clarity.- [ ] Code changes are covered by tests

• [ ] Code changes have been evaluated for compatibility/integration with TrendAnalysis
• [ ] New functions added to __init__.py
• API.rst is up to date, along with other sphinx docs pages
• [ ] Example notebooks are rerun and differences in results scrutinized
• Updated changelog

[codecov-commenter]

Codecov Report

❌ Patch coverage is 95.45455% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 96.14%. Comparing base (5c674a9) to head (fd434f9).

Files with missing lines	Patch %	Lines
rdtools/normalization.py	93.33%	2 Missing ⚠️

Additional details and impacted files

@@               Coverage Diff               @@
##           development     #475      +/-   ##
===============================================
- Coverage        96.22%   96.14%   -0.09%     
===============================================
  Files               12       12              
  Lines             2279     2280       +1     
===============================================
- Hits              2193     2192       -1     
- Misses              86       88       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[martin-springer]

@mdeceglie - Commit 87d2b63 simplified the tz handling. Tests are already extensive with tz-naive and tz-aware unit tests. All are passing. The simplification gets us closer to the original implementation.