sphenix-tracking-l AT lists.bnl.gov

Subject: sPHENIX tracking discussion

List archive

Re: [Sphenix-tracking-l] Tracking quality with Kalman Pattern recognition and macro from pull request 60

From: Haiwang Yu <yuhw.pku AT gmail.com>
To: "Tarafdar, Sourav" <sourav.tarafdar AT vanderbilt.edu>, "Frawley, Anthony" <afrawley AT fsu.edu>, Christof Roland <christof.roland AT cern.ch>, Jin Huang <jhuang AT bnl.gov>, Carlos Perez <carlos.perezlara AT stonybrook.edu>
Cc: "sphenix-tracking-l AT lists.bnl.gov" <sphenix-tracking-l AT lists.bnl.gov>
Subject: Re: [Sphenix-tracking-l] Tracking quality with Kalman Pattern recognition and macro from pull request 60
Date: Tue, 30 May 2017 16:35:06 +0000

Hi Sourav,

Thanks a lot for testing the new module and new macro.

I compared your results with Tony's. And made some plots as attached.

slide 2, 3: both Tony and Sourav turned on TPC charge distortion, which I didn't test at all. Great to know the code is still making tracks with the distorted TPC clusters
slide 4: there are non-zero pT offset in Tony's results, but not in Sourav's, which is very interesting. For my first impression was Sourav was using some older local version (slide 5, the cluster assoc. eff. vs layer in Sourav's looks like using an older version.)
slide 5: tracking eff. without momentum matching cut looks nicer, grass caused by low inner tracker assoc. eff. could be the reason why there is another 10% drop when using momentum matching
slide 6: despite lower inner tracker cluster assoc. eff. Sourav's DCA distribution looks similar with Tony's, which is also interesting.

Any how, I will do more tests myself to verify those guesses.

Cheers,

Haiwang

On Mon, May 29, 2017 at 10:43 PM Tarafdar, Sourav <sourav.tarafdar AT vanderbilt.edu> wrote:

Hello Tony and Haiwang,

I was recently looking at the tracking performance using the macro from pull request 60. I don’t really see the linear response in the pT resolution plot for both pure pions and Hijing embedded pions as shown by Tony during tracking meeting last week. Also the track efficiency looks bit different. It goes > 90% for pure pions until 20-25 GeV and then somehow starts decreasing. This feature of decreasing efficiency at high pT is more prominent for Hijing embedded pions. I applied only 4 sigma matching cut for calculating the efficiency. The DCA resolution looks far too better here for pT < 5 GeV as compared to the plots shown on Friday. I have listed the location of the relevant files in the attachment and also the quality related plots. Any thoughts ?

Thanks,
-Sourav

Attachment: KalmanPatRec_2017-05-30_simulation_meeting.pdf
Description: Adobe PDF document

Re: [Sphenix-tracking-l] Tracking quality with Kalman Pattern recognition and macro from pull request 60, Haiwang Yu, 05/30/2017