sPHENIX tracking discussion

["Yasuyuki Akiba" <akiba AT rcf.rhic.bnl.gov>]

Dear Tony,

 

Because of the echo in Bluejeans, I am afraid that what I said in the meeting was not well heard. So I will write it here.

 

The plot below is electron DCA distribution and its breakdown after unfolding. The bàe contribution is shown in light blue, and you can see a flat DCA background at the level of 3 orders of magnitude below the peak. The flat background is cause by that one or more hits in VTX are associated with the track. The b signal is just above the flat background. So if the background level is significantly higher, you cannot tag b.

 

The level of the grass is very important for heavy flavor tagging. From my experience of VTX analysis, I can tell you that it is relatively trivial to achieve a good DCA resolution in the main gaussian peak of the DCA distribution. But no matter how good the DCA resolution of the gaussian peak, you cannot do heavy flavor tagging or b/c separation if you have high level of flat, random background. This background is caused by the mis-association of 1 or 2 noise hits or mis-association of the outer tracker and the VTX tracklet.

 

The point I want to make is that the level of the flat background in the DCA distribution is the key performance parameter in a vertex tracker. The evaluator plot currently does not have this key performance plot. It should be added before the review.

 

Sincerely yours,

              Y. Akiba

From: sphenix-tracking-l-bounces AT lists.bnl.gov [mailto:sphenix-tracking-l-bounces AT lists.bnl.gov] On Behalf Of Yasuyuki Akiba
Sent: Friday, August 12, 2016 8:48 AM
To: 'Michael P. McCumber' <mccumber AT bnl.gov>
Cc: sphenix-tracking-l AT lists.bnl.gov
Subject: Re: [Sphenix-tracking-l] sPHENIX tracking meeting 9:00 am Friday August 12

 

Dear Mike,

 

Thank you for the prompt reply and clarification.

 

So where is the DCA distribution from the evaluator?

 

If at least 1 noise hit is picked up, this should cause a large DCA for the track. So the DCA distribution should have a grass of wide DCA tail of 5% of all tracks.

 

Sincerely yours,

              Y. Akiba

 

From: michael.p.mccumber AT gmail.com [mailto:michael.p.mccumber AT gmail.com] On Behalf Of Michael P. McCumber
Sent: Friday, August 12, 2016 8:42 AM
To: Yasuyuki Akiba <akiba AT rcf.rhic.bnl.gov>
Cc: sphenix-tracking-l AT lists.bnl.gov
Subject: Re: [Sphenix-tracking-l] sPHENIX tracking meeting 9:00 am Friday August 12

 

Hello Yasuyuki,

 

The inner purity plot does indicate that 5% of the time a noise hit is picked up in the inner 3 layers. However the "vertex residual" plot is a collision vertex reconstruction resolution, it is not a plot of the track DCA distribution.

 

Mike

--

Michael P. McCumber, PhD
Los Alamos National Laboratory
505-709-8161

 

On Fri, Aug 12, 2016 at 4:33 AM, Yasuyuki Akiba <akiba AT rcf.rhic.bnl.gov> wrote:

Dear Mike,

 

I looked your slides posted in the CDS agenda and have a question.

 

In slide 5, you have

What is the meaning of “inner finding efficiency” and “inner purity”? What is the exact definition?

From the word “inner”, I suppose this is plot is for 3 layers of MAPs.

So I guess “exact match efficiency (purity)” means the efficiency (purity) that the TPC track is associated with the correct hits in all of the 3 MAPs layers. Is this correct?

 

If this is correct, I have a related question n the vertex residual.

 

According to the inner purity plot, the probability of exact match is about 95%. This means 5% of the tracks are wrongly associated to at least 1 hits. Then these 5% track should have “wrong DCA” measurement. But there is no grasses of the DCA distribution. This is very strange.

 

I suspect that this could be artifact due to that all tracks are coming from DCA=0. There is no track that has a real large DCA. Since all track segment in the 3 layers of MAPs is a straight line coming from the event vertex, their DCA values is very close to zero. So, even if a TPC track is associated with a wrong hit in one or more of MAPs layer, the DCA can be still zero.

 

Sincerely yours,

              Y. Akiba

 

 

 

From: sphenix-tracking-l-bounces AT lists.bnl.gov [mailto:sphenix-tracking-l-bounces AT lists.bnl.gov] On Behalf Of Frawley, Anthony
Sent: Wednesday, August 10, 2016 7:45 PM
To: sphenix-tracking-l AT lists.bnl.gov
Subject: [Sphenix-tracking-l] sPHENIX tracking meeting 9:00 am Friday August 12

 

Hi All,

We will have an sPHENIX tracking meeting on Friday, August 12 at 9:00 am ET. The agenda page is:

 

https://indico.bnl.gov/conferenceDisplay.py?confId=2325

 

The focus of the meeting will be progress and plans for the tracking review.

 

Best regards

Tony

_______________________________________________
Sphenix-tracking-l mailing list
Sphenix-tracking-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-tracking-l