Difference between revisions of "Mattione Kaon PID Studies eta1Prime(2300)"

Revision as of 22:56, 14 September 2012

Reaction

• γ p → η'₁(2300), p
  • η'₁(2300) → K*(892)⁰, K⁰_bar
    • K*(892)⁰ → K⁺, π^-
      • K⁰_bar → π⁺, π^-
• Final State: p, K⁺, π^-, π⁺, π^-
• Events generated with: Γ_η'1(2300) = 200 MeV/c², t-slope = 5.0, 50.0 < Vertex-Z (cm) < 80.0

Analysis Overview

• For each q = +/- particle in the reaction, all detected q = +/- tracks are tested as to whether or not they are that particle
  • e.g. If a q = + track has PID FOM > 1% for both the π⁺ & K⁺: allow both combinations (DO NOT take the best FOM as true!)
• Maximum of 8 tracks per event (to cut down on memory spikes).
• The analysis (cuts/histing/kinfits) is then performed on each possible track combination (events are only cut if every one of their track combinations is cut):
  • Required 45.0 < Track-Vertex-Z (cm) < 85.0
  • EXTREMELY loose mass cuts around each mass peak (typically at the histogram edges): saves time by reducing # of kinematic fits to perform
  • Require the combined FOM for PID (χ² and NDF of all tracks summed together) be > 1%
  • Where kinematic fit: Require that the confidence level from the kinematic fit be > 1%
  • Mass cuts as illustrated.

Mass Peaks & Kinematic Fitting

genr8 η'₁(2300) Signal

• Kinematic Fit Confidence Level (Zoomed in; spike at 0 confidence is well off the top of the plot)

• K⁰ Invariant Mass (Before/after KinFit where present)

• Missing Mass & Missing Mass Squared (Before/after KinFit where present)

• K*(892)⁰ Invariant Mass (Before/after KinFit where present)

• η'₁(2300) Invariant Mass (Before/after KinFit where present)

genr8 η'₁(2300) Background

• Generate pions instead of kaons, shift generated mass peaks so that when the kaon mass is assigned to the kaon candidates (pions), the mass peaks shift to the original signal masses.
  • Note that it is impossible to get every peak to line up without generated a recoil proton with a bogus mass. Therefore, the missing mass peak is slightly off.

• K⁰ Invariant Mass (Before/after KinFit where present)

• Missing Mass & Missing Mass Squared (Before/after KinFit where present)

• K*(892)⁰ Invariant Mass (Before/after KinFit where present)

• η'₁(2300) Invariant Mass (Before/after KinFit where present)

bggen Background

• K⁰ Invariant Mass (Before/after KinFit where present)

• Missing Mass & Missing Mass Squared (Before/after KinFit where present)

• K*(892)⁰ Invariant Mass (Before/after KinFit where present)

• η'₁(2300) Invariant Mass (Before/after KinFit where present)

Cherenkov Detector Comparison: bggen Background

• P vs. θ of π⁺'s as K⁺ candidates ("All" means all post-analysis cuts, pre-Cherenkov cuts)
  • π^-'s as K^- not shown: identical due to symmetry.

Yields %'s

• ALL yield %'s are determined in the signal region (track combinations with invariant mass outside of the signal region are ignored)

genr8 η'₁(2300) Signal Events

•  % of generated events remaining after all cuts (including a final cut requiring that all tracks have the correct THROWN PID)

•  % of BAD reconstructed combos remaining (BAD combo: a combo where at least one track has the wrong THROWN PID)

genr8 η'₁(2300) Background Events

•  % of reconstructed combos remaining

bggen Background Events

•  % of reconstructed combos remaining