Hello All,

I have updated the bunch selection information page, and I think the goal
is very close.
   http://ccjsun.riken.go.jp/~jinnai3/polarimeter/bunch_sel/ 

(I needed to make decisions on some values such as threshold etc to 
 proceed. Please consider them as a tentative numbers, and let me know 
 if you feel they are not proper.)

o In order to decide the Chisq/ndf cut threshold, asymmetry plots are 
  made to compare the difference.
  - lets take a look at physics asymmetry and luminosity asymmetry 
    plots in 2.b.  
    * Luminosity asymmetry plots -- Chisq/ndf = 3 cut is not acceptable 
      because they are distributed up to 20%, chisq/ndf=5 is acceptable.
    * Physics asymmetry plots -- if we look at yellow Y-components, the 
      tight cut is preferable.
  - Chisq/ndf = 5 is adopted for the moment 

o Files containing the bunch lists with the keep/discard flag are 
  prepared in 2.d. Flags are assigned in accordance with the cut 
  conditions which have been discussed previously.

o In 2.e, the criteria to throwing out the CNI runs is decided from 
  the plots.    

  and in 2.f, tentative list of FILLs and RUNs should (will) be 
  discarded in the spin analysis is shown.

Misc. The emails so far are put together at the end of the page.  

----------------------------------------------
To do,    
    - Determine the bunch selection rules after the comments and 
      suggestions from the receivers of this email.  

    - Finalize the bunch selection with the input from Phenix.
      (we have already got the input from STAR)  

    - Finalize the asymmetry (polarization) values.
      Which format are they preferred, fill by fill, or run by 
      run(CNI)? or both?

Remarks,
    I have to apologize for the delay of the work and even now the 
    finally values are not available. The comments and discussions 
    are helpful and necessary to accelerate this work.

with best regards,
    Osamu

P.S. I have added Brendan and Takahiro in the list