Added pooled scoring and robust aggregation

[jigyasaba]

** Summary:**

This PR resolves the existing TODO regarding score_ by exposing a pooled global metric for both GWLogisticRegression and GWLinearRegression.

Motivation:
While local models were fitted correctly in the base classes, a global summary metric consistent with the scikit-learn API (score_) was not explicitly provided.

Geographically Weighted (GW) models produce location-specific results, but users expect an overall performance summary comparable to standard sklearn estimators. This PR introduces a pooled metric based on focal predictions.

Implementation:

GWLogisticRegression

• score_ defined as pooled mean accuracy of focal predictions.
• NaN focal predictions are masked to avoid invalid aggregation.
• Added safeguards when concatenating pooled arrays to prevent shape issues when local models are skipped.

GWLinearRegression

• score_ defined as pooled R² computed from focal predictions.
• Guard added for division-by-zero when total variance (ss_tot) equals zero.
  The design uses focal predictions rather than weighted aggregation, ensuring consistency with the interpretation of GW estimators as a spatial ensemble of local models.

Impact:

• No API-breaking changes.
• Backward compatible.
• Aligns GW estimators more closely with sklearn estimator conventions.
• Improves stability when some neighborhoods are skipped due to invariance or min_proportion.

All existing tests pass locally.
Open to feedback and suggestions.

[codecov]

Codecov Report

❌ Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 92.95%. Comparing base (9b62eef) to head (f0a5df8).

Files with missing lines	Patch %	Lines
gwlearn/linear_model.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #97      +/-   ##
==========================================
- Coverage   92.96%   92.95%   -0.02%     
==========================================
  Files           6        6              
  Lines         825      837      +12     
==========================================
+ Hits          767      778      +11     
- Misses         58       59       +1     

☔ 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.

[martinfleis]

Duplicate of #76. Also, see the discussion in there.