jin huang, xin qian for transversity analysis meeting may 3, 2010

12
Final Story on Cross Check between Red Team and Blue Team Results Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Upload: christine-lawrence

Post on 17-Jan-2018

219 views

Category:

Documents


0 download

DESCRIPTION

 simpler to check DSA, which we start with  Using same method for both team ◦ Divide phi_h – phi_S to 100 angular bins ◦ For each bin, calculate raw asymmetry  A raw = (N + - N - )/ (N + + N - ) ◦ Then fit the DSA modulation  We found ◦ The difference remains ◦ Each team’s result is consistent with its own previous one  Therefore ◦ Good News: each team is self consistent, not likely a coding error ◦ Difference from low level of the analysis Transversity Analysis Meeting Jin Huang 3

TRANSCRIPT

Page 1: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Final Story on Cross Check between Red Team and Blue

Team ResultsJin Huang, Xin Qian

For Transversity Analysis MeetingMay 3, 2010

Page 2: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 2Transversity Analysis Meeting

From Last Meeting: Difference remains w/ run cut sync After run list/cuts/variable adjustment

◦ DSA Asymmetry is very consistent◦ Angular modulation became more consistent (not totally)◦ The remaining difference is investigated during last week

Page 3: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 3

simpler to check DSA, which we start with Using same method for both team

◦ Divide phi_h – phi_S to 100 angular bins◦ For each bin, calculate raw asymmetry

Araw = (N+ - N-)/ (N+ + N-)◦ Then fit the DSA modulation

We found◦ The difference remains◦ Each team’s result is consistent with its own previous one

Therefore◦ Good News: each team is self consistent, not likely a

coding error◦ Difference from low level of the analysis

Transversity Analysis Meeting

We tried … and found …

Page 4: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 4

The phi_S, theta_S data had problem

Transversity Analysis Meeting

Further check we found

Page 5: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 5

Both team are using their own copy of root files developed from original skim ones.

We found there is a problem on phi_S -> phi_S + pi affecting all transverse root files last October

A fix is applied blue team files in December◦ It was correct

Another patch was generated for my copy of root files in January◦ However, ~10% of run was not corrected in the

patch

Transversity Analysis Meeting

History of the problem

Page 6: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 6

Xin generated an updated patch for my copy of data Multiple Check event by event Confident no further problem on spin angles

Transversity Analysis Meeting

What now?

Page 7: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 7

A_LT modulation with polarization correction

Transversity Analysis Meeting

DSA agree well now

Page 8: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 8Transversity Analysis Meeting

SSA difference

•Red: MLE Final•Blue: Blue team Final

•Difference could be40% of σ for 1 term fit•2 term fit is more consistent:

Page 9: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 9

The event list is very close Method

1. Combine Spin States: Local Pair VS MLE Results have difference We know MLE is more sensitive to yield drifts Believe Local Pair method is better for Pion analysis

2. Extract Angular Modulation: Angular Fit VS MLE Believe to be equivalent @ large stat. (ex. (e’pi))

Indirectly tested on DSA Tested for both 1 term 1D fit and 2 term 2D fit

Xin showed that at low stat. (ex. (e’K)), MLE have better statistic precision

Transversity Analysis Meeting

SSA difference is understood

Page 10: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 10

Raw SSA Based on very similar data set

Difference believed to be from different way combining local spin states

Transversity Analysis Meeting

DIFFERENCE in combining spin states

Page 11: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 11

Indirectly checked by fitting sin(phi_h±phi_S) on DSA:

Transversity Analysis Meeting

CONSISTENCY in Angular Modulation Extraction

Page 12: Jin Huang, Xin Qian For Transversity Analysis Meeting May 3, 2010

Jin Huang <[email protected]> 12

Difference on Method ◦ Blue team Fit method is better for Pion SSA

analysis◦ MLE method is better for Kaon SSA/DSA analysis◦ Both MLE/Fit methods OK for DSA pion analysis

Difference on Cut/Run List ◦ Remove Extra 17 Problematic Runs (L1A/DMA/…

problem) identified by blue team◦ 8 Runs with low LT remain off from MLE run list.

OK for blue team method (local pair with in these runs)

◦ Remove “LHRS pion rejecter response>0” cut 3~4% more event in statistic

◦ Other difference on Cut/Variable use are negligible

Transversity Analysis Meeting

Difference understood &Suggestion on addressing them