Data reduction

These pages document the reduction layer between the raw detector histograms and the asymmetry curve: detector balancing (alpha), background handling, histogram binning, time-zero determination, detector exclusion and multi-period mapping. Every quantity here feeds all downstream analysis — a weak alpha estimate or a missed background degrades every fit made from the data — so each page states explicitly when its method applies and when it does not.

• Alpha calibration
  • Choosing a method
  • Corrections and the estimate
  • Diamagnetic method
  • General method
  • Count ratio
  • Count-fit α
  • Uncertainties
• Background correction
  • Choosing a mode
  • Range average (pre-t0)
  • Tail fit (late-time)
  • Background run
• The background ladder: which stage removes what
  • The ladder at a glance
  • Stage 1 — the grouping background mode (pre-asymmetry, pre-FFT)
  • Stage 2 — the count-fit background nuisance (in the fit, on raw counts)
  • Stage 3 — the σ-clip baseline (post-FFT, display)
  • Stage 4 — MaxEnt SpecBG (post-reconstruction, display, ZF/LF)
  • Why they coexist
  • See also
• Histogram binning
  • Fixed
  • Variable
  • Constant error
  • Rebinning programmatically
  • Notes
• Time-zero search
• Detector exclusion
  • Identifying a bad detector
• Period mapping

All of the corrections described here are applied at reduction time and recorded in the grouping settings: raw histograms are never modified, and a project re-opened later reproduces the same reduction from the same recorded choices.