Add comprehensive statistical inference for TRF

[Hugo-W]

Add comprehensive statistical inference for TRF

Description

Current TRF computation focuses on coefficient estimation. Need to add comprehensive statistical inference capabilities for naturalistic data analysis.

Tasks

• Create TRFAnalyzer class (or module) for statistical analysis
• Implement:
  • Permutation testing with cluster-based correction for temporal correlations
  • Bootstrap confidence intervals on TRF coefficients
  • Cross-subject consistency measures (spatial reliability)
  • Time-frequency decomposition of significance
• Handle violations of standard statistical assumptions in naturalistic data (temporal dependencies)
• Integrate with existing TRFEstimator

Acceptance Criteria

• Comprehensive statistical methods for TRF analysis
• Proper handling of naturalistic data characteristics
• Integrates cleanly with existing TRF workflow

Priority

Low

Dependencies

• Issue Move root solver.py into pyeeg/solvers.py, consolidate, and test #8 (Create models/ subpackage) recommended first
• Issue Replace placeholder tests in test_utils.py #9 (Solver Pattern) recommended first

Epic

New Features

Related

From external analysis: Advanced Statistical Inference Module