Jin Huang, also for Xin Qian For Transversity Analysis Meeting Feb 20, JLab
There are difference between red and blue team results ◦ ~70% points consist ◦ For problem points, discrepancy could be as large as 1 sigma ◦ Exist for both SSA and DSA, Asymmetry and Modulation Xin and I have been hunting these discrepancy for the last week ◦ Cuts/Run Lists ◦ Method Transversity Analysis Meeting Jin Huang 2
Transversity Analysis Meeting Jin Huang 3 ~1σ Discrepancy:
Transversity Analysis Meeting Jin Huang 4 ~1σ Discrepancy:
1. Run List 2. Cuts/Variable choice 3. Method Difference A.Combining Spin State Blue: Super Local Pair Red: One step MLE B.Form angular modulation Blue: Angular Bin & Fit Red: One step MLE Transversity Analysis Meeting Jin Huang 5
Difference: ◦ Blue team excluded 20 more runs: Yield out 3σ of local average ◦ Red team excluded 3 more runs: w/ Low Life time due to PS mistakes or RICH tests For following study: ◦ Use blue team run list for red team analysis Transversity Analysis Meeting Jin Huang 6
Found 5 Difference in cuts / variable usage ◦ Red team cut on pion rejecter E/p>0 (e’pi) events 3~4% of total coincidence events most of event difference ◦ pT cut, BB_ts_ps_e_c[i]>400 or BB_ts_ps_e[i]>400 ◦ Variable choice: Beam curent & edtp choice For following study: ◦ Use blue team cuts/variable choices Transversity Analysis Meeting Jin Huang 7
After the synchronization: ◦ Event sample difference no more than ◦ In the data set generating above plot Red team excluded events during spin flip while blue team cut them out in later stage ◦ All but one event in Red Team sample is in Blue team’s ◦ Red team excluded more events around spin flips (2e-4 difference) Transversity Analysis Meeting Jin Huang 8
Double spin asymmetry agree well now Transversity Analysis Meeting Jin Huang 9
Better after run/cuts sync Some difference remains for Single Spin Asymmetries Method Difference ◦ MLE method is known more biased by yield drift than super local pair method in SSA ◦ Not problem the case for DSA Transversity Analysis Meeting Jin Huang 10
Transversity Analysis Meeting Jin Huang 11 After run list/cuts/variable adjustment ◦ As shown asymmetry is consistent for DSA. ◦ Angular modulation also became more consistent ◦ Some difference remains: under invesitagtions
Transversity Analysis Meeting Jin Huang 12 After run list/cuts/variable adjustment ◦ As shown asymmetry is NOT all consistent for SSA Could be as large as 50% of σ for low x for pi- ◦ Angular modulation became more consistent in general ◦ Some difference remains 2 term fitting more consistent than 1 term fitting See next two slides for plots
Transversity Analysis Meeting Jin Huang 13 Asymmetry also improved equal much for pi+ high x and low x bin: ~0.5σ Discrepancy on Asymmetry. Also equal Discrepancy on modulation:
Transversity Analysis Meeting Jin Huang 14 Points improved: >0.5σ Discrepancy:
Run List/Cuts/Variable difference ◦ Responsible for most difference ◦ not a problem anymore Method Difference ◦ Combining Spin State not a problem for DSA Local pair method has been expected to be better for SSA ◦ Form angular modulation Under investigation (red team running angular fitting) (expected) responsible for difference in DSA. Could also involve SSA discrepency Transversity Analysis Meeting Jin Huang 15