Analysis Strategy

From Hchkswiki
Jump to: navigation, search

Analysis Strategy

Based on the meeting on 18th Nov 2009 at F324.

We agreed that:

All analysis will base on the same first level ntuple/root file.
The replay code will be controlled by subversion.
Wiki and new ELog (to be set up) will be fully used. 
We will have periodical TV-conferences to share analysis information.

Here is a schematic flow of the analysis. Analysis-flow.png

Here is work list based on Liguang's draft (Media:analysis work list.doc). Responsible person for each task will be assigned soon.

  • Produce 1st level ntuple/root to be placed at the shared directory
  1) 1st level ntuple structure
    HAPPEX dithering, HALLCp, 9th Dipole NMR, Hall C Arc BPM’s (3C07, 3C12, and 3C17) must be included.
  2) Establish common definition of coordinate, kinematic definition.
  3) “standard” initial cuts and conditions (should not depends on tracking)
  4) separate ntuple can be made based on trigger (CP0, single, coin etc...)
  5) scaler information including integral charge
  6) Remove orvious background like beta=1 particle in the HKS arm
  7) Corrupted data should be adequately fixed for the second level analysis 

  • Finalize the standard replay code
  1) Adequate treatment of the multihit events
  2) Clean up of the tracking code (We know the existing code have lots of memory leaks)
  3) Finalize EDC1 code by using EDC2 information
  4) Optimize HDC tracking code
  5) HDC/EDC X-T calibration for various trigger conditions
  • Establish standard GEANT simulation
  1) Update TOSCA magnetic field with the final HP/NMR/Current set. Should be checked with field map result.
  2) Update GEANT model with the TOSCA field and the geometrical information from the survey.
  3) Coordinate/kinematic definition should be correctly associated to the replay code.
  4) Generate initial angular/momentum matrices which will serve as the "reference".
  • Optimize KID parameters
  1) HTF, EHOD timing should be adjusted (slewing corrections, time zero adjustment) for each run condition. Rate was so different depending on targets and beam current. 
  2) Aerogel Chenrenkov parameter tuning, kaon loss, pion rejection efficiencies should be estimated with CP0 data.
  3) Water Chenrenkov parameter tuning, kaon loss, proton rejection efficiencies should be estimated with CP0 data.
  4) Lucite Cherenkov
  5) Improve KID selection method 
   5.5) Study about sophisticated selection method like : Likelihood cut, neuronetwork etc.
  • H2O data analysis with 0 and +-1MeV energy offset : Optics is fixed
  1) Using mass shifts of the Lambda peak, beam energy shifts should studied to check consistency
  2) Check HALLCp and 9th dipole NMR   
  3) Represent the beam energy as a function of BPMs and obtain the beam energy correction formula.
  • H2O and C data analysis with 0 and +20MeV energy offset : Kinematics is fixed
  1) With the optimized kinematics, Lambda and B12L(gs) peaks are used to check linearity of the mass scale. 
  2) If S/N ratio is good enough, optics tune can be processed
  • Target energy loss/straggling, radiation tail shape study with simulation codes
  1) e, e' and K+ energy loss and straggling, radiation tail shape
  2) GEANT4 simulation
  3) SIMC 
  • Optics tuning
  1) Sieve slit data are used for angular matrices
  2) Lambda, Sigma peaks of CH2 data and Lambda peak of H2O, B12L(gs) peak are used for momentum matrics
  • Missing mass analysis