Skip to content

Reference

This section contains the detailed API documentation for the public ffTRF surface.

Start Here

  • Read TRF Estimator for the main class and its shared parameter semantics.
  • Read Result Containers for the objects returned by the frequency-resolved, transfer-function, diagnostic, and significance helpers.
  • Read Metrics for built-in scoring functions and custom-metric requirements.
  • Read Segment Settings Helper for the public rule-of-thumb helper that suggests segment_duration, overlap, and window.
  • Read Preprocessing Helpers for small signal-preparation utilities shipped with the package.

Public API Summary

  • TRF: train, predict, score, run score-level or refit-based significance tests, inspect, plot, diagnose, save, and load models
  • TRFDiagnostics / CrossSpectralDiagnostics: observed-vs-predicted spectral diagnostics
  • PermutationTestResult: held-out score significance against surrogate nulls
  • FrequencyResolvedWeights: frequency-by-lag decomposition of the kernel
  • TimeFrequencyPower: spectrogram-like power view of the kernel
  • TransferFunctionComponents: magnitude, phase, and group delay for one transfer-function slice
  • pearsonr, r2_score, explained_variance_score, neg_mse: built-in metrics
  • suggest_segment_settings: helper for choosing practical segment and window defaults for the standard estimator
  • half_wave_rectify, resample_signal, inverse_variance_weights: small preprocessing utilities