Configuration Deep Dive

This section is a standalone, human-first guide to how PQC configuration maps to detector behavior and statistical interpretation.

• Expanded standalone explanation of the PQC configuration
  • 1. What this pipeline is for
  • 2. What the run configuration means
  • 3. The logic of the whole detector stack
  • 4. Bad-point detector: OU model, FDR control, and per-day marking
  • 5. Transient exponential-recovery detector (tr_cfg)
  • 6. Exponential dip detector (dip_cfg)
  • 7. Metadata merge (merge_cfg)
  • 8. Derived features (feature_cfg)
  • 9. Feature-structure diagnostics and detrending (struct_cfg)
  • 10. Achromatic step detector (step_cfg)
  • 11. Chromatic step detector (dm_cfg)
  • 12. Robust outlier detector (robust_cfg)
  • 13. Detector preprocessing (preproc_cfg) and hard sigma gate (gate_cfg)
  • 14. Solar-event detector (solar_cfg)
  • 15. Orbital-phase cut (orbital_cfg)
  • 16. Eclipse detector (eclipse_cfg)
  • 17. Gaussian bump detector (bump_cfg)
  • 18. Glitch detector (glitch_cfg)
  • 19. Why some detectors are backend-specific and others are global
  • 20. How the significance thresholds work in practice
  • 21. The semantic rule that changes everything
  • 22. A fully plain-English summary of the whole config
  • 23. Suggested references for readers
  • 24. A shorter “TL;DR” version
  • References