sPHENIX tracking discussion

[Hugo Pereira Da Costa <hugo.pereira-da-costa AT cea.fr>]

More bellow:

On 11/16/21 12:07, Osborn, Joe wrote:

Hi Hugo,

 

Thanks for the responses, I think they make sense. The idea of implementing event-by-event fluctuations within the process_event of the initial module is a good one.

 

> My understanding is that the CA seeder uses 3D coordinates anyway. Inside the seeder, one would get the 2D + surface ID from the cluster, convert to 3D, apply the correction and use in the CA algorithm, instead of the uncorrected 3D position. Does that make sense ? In more details, I plan to have a DistortionCorrection::get_corrected_postion( TrkrCluster* ), that could be used inside the CASeeder code. (note that I have not yet looked into the CA Seeder code).

Okay, that makes sense. I think that is a good design as it will allow for simple access to any “corrected” cluster position given the original distorted measurement anywhere in the workflow (as you had suggested).

 

My only wonder is if we will actually need this in the seeder, since the CA seeding seems to perform well enough on the distorted clusters other than getting the track parameters quite wrong.

Maybe this touches the discussion from this morning on whether or not we include this in the circle fit code, which with the initially corrected clusters will presumably provide a better TPC seed pointing resolution and allow us to match to silicon tracks.

 

    I was wondering the same too. It might be enough to implement it in
    Tony's circle fit, to get proper parameters, and not in the seeder.
    Still, if the correction lookup time is not a problem, then it could
    not hurt to have it already in the seeder ... To be decided once the
    code is in place, I guess.

On the CA seeder working properly with fully distorted clusters, I had a question on your slides from yesterday, which I forgot to ask: are the performances you've shown (number of seeds per events, number of layers per seed, w/ w/o distortions) directly after the PHCASeeder, or after the KFPropagator ? (if I understand right, the later is used to make the seeds longer, i.e. add more clusters, but might rely more on the track parameters)

In any case I think I understand now, and don’t have any comments on your suggested design.

cool !

 

Joe

 

---------------------------   Joe Osborn, Ph.D. Associate Research Scientist Oak Ridge National Laboratory osbornjd AT ornl.gov (859)-433-8738

 

 

From: Hugo Pereira Da Costa <hugo.pereira-da-costa AT cea.fr>
Date: Tuesday, November 16, 2021 at 1:16 PM
To: Osborn, Joe <osbornjd AT ornl.gov>, sphenix-calibration-l AT lists.bnl.gov <sphenix-calibration-l AT lists.bnl.gov>, sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>
Subject: Re: [Sphenix-calibration-l] [EXTERNAL] Re: TPC Distortions meeting, Tuesday, 11AM Eastern

Also: thanks for the offer for help. Now really this is not much work, (provided we agree on a path forward): mostly moving around code that already exists in offline/packages/tpc/TpcSpaceChargeCorrection.cc.

Hugo

 

 

On 11/16/21 11:12, Hugo Pereira Da Costa wrote:

Hi Joe,

thanks for the comments. Answers below:

On 11/16/21 11:04, Osborn, Joe wrote:

Hi Hugo,

 

Thanks for the set of slides – here are some of my initial thoughts. BTW I am happy to help with this also, if it would be useful to divide it up. Just let me know if I can help.

 

I guess this would only account for static and average distortions if they are loaded at InitRun. Is my understanding correct? How would we handle the event-by-event fluctuations, would this be a histogram (or whatever structure) that gets updated in another module and could be stored as a part of these objects?

That is a good question. It is not clear to me how per-event maps will be handled anyway. On the fly ? in a first pass and saved into a file, as a Tree of histograms ?

For the former, I can imagine having the LoadDistortionCorrection module also have a process_event hook that take care of updating what's in the node tree on a per event basis. (that can then be accessed by the other modules)

For the latter it is only a matter of loading said tree of histograms in InitRun.

But yes, for now I was planing to only implement static+average, which is the only thing well defined (and currently implemeneted anyway), and see later how to evolve that to also include per event corrections. Does that make sense ?

  For TpcDistortionCorrection, we discussed yesterday/today that in order to apply a correction to a cluster we will already need to know what track it is on and have the track projection from the silicon+TPOT. Maybe I am misunderstanding, but how would this interface with PHCASeeder? I would think we would need to also implement the machinery from the ClusterMover into this (which is maybe what you were intending).

My understanding is that the CA seeder uses 3D coordinates anyway. Inside the seeder, one would get the 2D + surface ID from the cluster, convert to 3D, apply the correction and use in the CA algorithm, instead of the uncorrected 3D position. Does that make sense ? In more details, I plan to have a DistortionCorrection::get_corrected_postion( TrkrCluster* ), that could be used inside the CASeeder code. (note that I have not yet looked into the CA Seeder code).

 

Hugo

 

 

Joe

 

 

---------------------------   Joe Osborn, Ph.D. Associate Research Scientist Oak Ridge National Laboratory osbornjd AT ornl.gov (859)-433-8738

 

 

From: sPHENIX-calibration-l <sphenix-calibration-l-bounces AT lists.bnl.gov> on behalf of Hugo Pereira Da Costa via sPHENIX-calibration-l <sphenix-calibration-l AT lists.bnl.gov>
Date: Tuesday, November 16, 2021 at 11:38 AM
To: sphenix-calibration-l AT lists.bnl.gov <sphenix-calibration-l AT lists.bnl.gov>, sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>
Subject: Re: [Sphenix-calibration-l] [EXTERNAL] Re: TPC Distortions meeting, Tuesday, 11AM Eastern

Hi Ross, Tony others

Unfortunately I have to leave early at today's TPC calibration meeting. I had a few slides to discuss what I would like to implement for "on demand" distortion correction, that can be used wherever necessary (e.g. CA seeding, cluster mover). Maybe we can discuss that offline instead ? The idea is to use similar code as is done with the ACTS surface map, and the local-to-global transformations.

Comments welcome

 

Hugo

 

 

 

On 11/16/21 09:03, Ross Corliss via sPHENIX-calibration-l wrote:

we can hold off a few, yeah.

==========
Dr. Ross Corliss

If it's work hours, and I'm not in a meeting, feel free to look for me in my virtual office:
https://stonybrook.zoom.us/my/rossoffice?pwd=ZmZ2SlRIMVFvUUJwbUkyOVNVTmE5QT09

 

On Nov 16, 2021, at 10:41 AM, Osborn, Joe <osbornjd AT ornl.gov> wrote:

 

I’m triple booked at 11, so if the tracking discussion can wait until 11:10 I would appreciate it – otherwise go ahead and I’ll join when I can.

 

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

 

Joe Osborn, Ph.D.

Associate Research Scientist

Oak Ridge National Laboratory

osbornjd AT ornl.gov

(859)-433-8738

 

 

From: sPHENIX-calibration-l <sphenix-calibration-l-bounces AT lists.bnl.gov> on behalf of Anthony Frawley via sPHENIX-calibration-l <sphenix-calibration-l AT lists.bnl.gov>
Date: Tuesday, November 16, 2021 at 10:26 AM
To: sphenix calibrations <sphenix-calibration-l AT lists.bnl.gov>, Ross Corliss <ross.corliss AT stonybrook.edu>
Subject: [EXTERNAL] Re: [Sphenix-calibration-l] TPC Distortions meeting, Tuesday, 11AM Eastern

Hello Ross,

 

We discussed tracking workflow with distortions yesterday in the tracking meeting.

 

I posted a couple of slides on that discussion.

 

Cheers

Tony

---

From: sPHENIX-calibration-l <sphenix-calibration-l-bounces AT lists.bnl.gov> on behalf of Ross Corliss via sPHENIX-calibration-l <sphenix-calibration-l AT lists.bnl.gov>
Sent: Monday, November 15, 2021 10:59 PM
To: sphenix calibrations <sphenix-calibration-l AT lists.bnl.gov>
Subject: [Sphenix-calibration-l] TPC Distortions meeting, Tuesday, 11AM Eastern

 

Dear Distortions Folks,

This is a reminder that the distortion meeting will take place tomorrow at the usual coordinates and time, 11AM EST.  I believe there are updates to discuss on both the tracking integration and readout simulation fronts, and other updates are encouraged, as always. 

https://stonybrook.zoom.us/j/94081538941?pwd=Y256N3ptUC9zV2M4cjNtUGgyWnJYdz09

https://indico.bnl.gov/event/13717/

-Ross

==========
Dr. Ross Corliss
Research Assistant Professor
Center for Frontiers in Nuclear Science
Stony Brook University
ross.corliss AT stonybrook.edu

_______________________________________________

sPHENIX-calibration-l mailing list

sPHENIX-calibration-l AT lists.bnl.gov

https://lists.bnl.gov/mailman/listinfo/sphenix-calibration-l