STAR 2013 W Test Production Analysis using Various Tracking and Vertex Finding Algorithms  

Production Summary

• Tracking Algorthms : official STI [additional HFT metrials incuded compared to run 13 official productio], EVAL STI* [*-some modification from Yuri] , EVAL STICA, STICA * [yuri's STICA] , STI_updated [ hits reuses for multiple tracks]
• Vertex Finading algorithms :         PPV_Default , PPV_W-mode, KFV
• Libraries :  Official SL16b ,  EVAL (created for tracking code development and evaluation) .
  • evals 1 = SL16b with STI + PPV_W [ additional HFT metrial's included]
    
  • evals 2 = EVAL :  STI* + PPV_W (*- Yuri's updates to STI)
  • evals 3 = EVAL : STI + KFV
  • evals 4 = EVAL : STICA +PPV_W  [similar to Yuri's prodction of St full W stream]
  • evals 5 = EVAL : STICA + KFV
  • evals 6 = SL16b_1 (updated SL16b) : STI_updated+ PPV_W 

documentation about STAR 2016 Library: https://drupal.star.bnl.gov/STAR/comp/sofi/soft-n-libs/library-release-history/2016-0#SL16b
Other libraryies :  https://drupal.star.bnl.gov/STAR/comp/sofi/soft-n-libs

---

---

Prodution Details

Lidia has produced 12.2 M events from period1 which is equal to total L2BW events from period 1 (dat 76-126) . Since we have removed few runs which failed our run 13 W QA we are been analyzing only 12.02 M events.

	label	Library	MuDst Locations	production log files	nEvents	W-Trees location
1	evals1	SL16b	/star/data23/reco/pp500_production_2013/ReversedFullField/evals1/2013/*/*	/star/rcf/prodlog/evals1/log/daq	12 M	/gpfs01/star/pwg_tasks/spin03/evals1
2	evals4	EVAL	/star/data19/reco/pp500_production_2013/ReversedFullField/evals2/2013/*/*	/star/rcf/prodlog/evals2/log/daq	12 M	/gpfs01/star/pwg_tasks/spin03/evals2
3	evals4	EVAL	/star/data40/reco/pp500_production_2013/ReversedFullField/evals3/2013/*/*	/star/rcf/prodlog/evals3/log/daq	12 M
4	evals4	EVAL	/star/data21/reco/pp500_production_2013/ReversedFullField/evals4/2013/*/*	/star/rcf/prodlog/evals4/log/daq	12 M	/gpfs01/star/pwg_tasks/spin03/evals4
5	evals5	EVAL	/star/data21/reco/pp500_production_2013/ReversedFullField/evals5/2013/*/*	/star/rcf/prodlog/evals5/log/daq	12 M
6	evals6	SL16b_1	/star/data19/reco/pp500_production_2013/ReversedFullField/evals6/2013/*/*	/star/rcf/prodlog/evals6/log/daq	12 M	/gpfs01/star/pwg_tasks/spin03/evals6

Run List

•  Run List used for the test productions : run-list-from-Lidia

• Run List used for the test production analysis : run-list-analysis

---

Meeting Notes - W test production analysis

According to Elke's suggestions Elke, Sal and run 13 W analysis group members (Devika, Matt, Bernd, Jinlong, Qinghua) and spin conveners agreed to discuss the test production analysis details in meetings.

03/29/16 :

• Production 1,2 are similar to official production, therfore should gives similar results. Priorities goes to Production 1,2 4 for W analysis. Then check for performance of KFV for W before test them.
• Need disk space.
• Plots that need to be looked in to

 
04/04/16 :

• Before making ratio plots we need to make sure same runs and same # of events were produced. For now get absolute number of W's / pb as. a function of ZDC rate

---------------------------------------------------------------------------------------------------------------------

---

Analysis Details

Technical Details / code issues: 

 

• For the libraries  "SL16b" and "stardev" , "eval" perhaps all the newer libraries we need to lord the class " StUeEvent " and this must be loaded before "StJetMaker " .And include ""StSpinPool/StUeEvent/StUeVertex.h" " in your Makers and in other codes where nedded.
• Compiling codes: For evals1 - use "starver SL16b" , and for evals2 -evals5 - use "starver eval"

Disk space where trees are stored:  /gpfs01/star/pwg_tasks/spin03/

BEMC -gain tables : For all the productions current DB BEMC (run 12 200 Gev calibration) gains were used.


The following plots were checked in comparison to official production. 

1. final W yields
2. golden W yields as a function of ZDC rate
3. W+/W- charge separation
4. signal to background ratio
5. W yields as a function of cuts
6. reconstruction of the W-recoil for A_N
7. All other imporatant plots for W-AL analysis.

Analysis Plots

For each and every set ,  total # of events which processed per run were checked and only runs which proccesed exactly the same # of events from each set during the w tree prduction were chosed for the comparison. So simply ratio plots can be checked between production 
# of Runs used in the comparison =  896 runs , 1.10 M events.

Test Production 1 : Evals 1

• Period 1 - W-analysis -pdf Movie
• Period 2

---

Test Production 2 : Evals 2

• Period 1 - W-analysis -pdf Movie
• Period 2

---

Test Production 4 : Evals 4

• Period 1 - W-analysis -pdf Movie
• Period 2

---

Comparison plots :

Evals 1 vs Evals2  
Comparioson plots : Click (here)

---

Evals 1 vs Evals4
Comparioson plots : Click (here)

---

Evals 1 vs Evals6
Comparioson plots : Click (here)

---

Evals 1 vs p14ia 
Comparioson plots : Click (here)

---

Evals 4 vs Yuri-P1
Comparioson plots : Click (here)

---

 
P14ia vs Yuri-P1 
Comparioson plots : Click (here)

---

P14ig vs Yuri-P2 
Comparioson plots : Click (here)

---

Compariso plots - ALL : Click (here)
Comparison plots -Updated
Comparison plots - spin PWG draft  ,  ver2 
Presentation to S&C meeting - ver1, ver2

W-Efficiency [Golden W / pb-1 vs ZDC] comparison in  STI  VS STICA

References