Home » Groups » STAR Protected

Sti Eloss Test After Eloss fix

Updated on Thu, 2014-08-28 11:22. Originally created by dongx on 2014-08-28 10:07.

 This post reports the findings of the Sti Eloss test using the Dmetric tool with the real data produced on Aug. 18 (before the fix) and Aug. 28 (after the fix).
All the plots show the difference between the case that StiHftC included and the case that all HFT geometries are disabled in Sti. Chain options are listed below:

Default chain (with StiHftC included - in P2014a)

bfc.C(200,"P2014a,btof,BEmcChkStat,Corr4,OSpaceZ2,OGridLeak3D,SCScalerCal,goptSCE100051,-hitfilt","/star/data03/daq/2014/047/15047100/st_physics_
15047100_raw_2500001.daq")

Chain with Sti HFT geometries all disabled)

bfc.C(200,"P2014a,btof,BEMCChkStat,Corr4,OSpaceZ2,OGridLeak3D,SCScalerCal,goptSCE100051,hitfilt,-PxlDb,-IstDb,-StiPxl,-PxlIt,-StiIst,-IstIT,-StiS
st,-StiSsd,-StiSvt,-SstIT","/star/data03/daq/2014/047/15047100/st_physics_15047100_raw_2500001.daq")
Here I did check the log file (searching "adding"),

/star/institutions/lbl_scratch/dongx/Sti_Test/0827/run_noHFT.log
one can see that no HFT geometries are loaded in Sti.

All production were run in DEV library.

  Aug. 18 (before fix) with a bit more statistics Aug. 28 (after fix)
 
 
 
 
     
     
     

A few comments:
1) There is no significant change before or after the energy loss fix for the single track over all momentum (energy). Here I mean not at the factor of 5-10, but I didn't check the precision within a factor of 2.
2) The scaled DCA (dca*pT^2) change with and without HFT geometries are within ~ 20 um*(GeV/c)^2. For 0.3 GeV/c track, the DCA change will be well within ~200 um. It is not so straightforward to understand the Dca global (mean) change at the order of 0.5-1mm in Jason's simulation study whether it comes from the Sti geometry.
3) The DEV chain as of now seems to hang at some event. The root4star didn't crash, but I didn't see the log come out anymore. It has been sitting there for hours. I didn't have enough time to investigate other details. That is also the reason the events produced after fix are smaller.
To reproduce this issue,
bfc.C(420,500,"P2014a,btof,BEmcChkStat,Corr4,OSpaceZ2,OGridLeak3D,SCScalerCal,goptSCE100051,-hitfilt","/star/data03/daq/2014/047/15047100/st_physics_15047100_raw_2500001.daq")


Groups: