sPHENIX discussion of physics

[Cameron Thomas Dean <cameron.dean AT cern.ch>]

Dear all,

Please find the minutes of Wednesday’s meeting below.

Cheers,
Jin, Marzia, Tony and Cameron

Introduction

 

Jin – New batch of c-cbar MC is generating with the INTT geometry fix. It will be fairly large and can be used to exercise the new DST formats. There will be an email when it is complete.

 

DST needs, HF-resonances, Tracks

 

Tony – Giving a track position at a vertex (PV) is tricky as you need to know which vertex the track comes from. It’s easier to extrapolate to the beamline. A vertex position is very tricky, for example a single particle event or for secondaries.

Jin – Very good point. Originally I had in mind that the position would be at the beamline and we would use Kalman filters to swi back and forth.

 

Joe – With regards to how we compare triggered and streamed data, there is some redundancy in the bunch crossing (BC), and we should have some association of which tracks belong to which vertex (a list of which tracks went in to the vertex estimate? – Cameron from memory)

Tony – The TPC track seed is the basic track and is associated to as many silicon seeds as fits the hypothesis. The silicon seeds have a unique BC but we don’t have a unique PV association. There is a non-zero possibility of having more than one PV per BC.

 

Xin – Under the track object, we have the number of clusters. Can we save the hit map to know which layers were hit?

Joe – You can access this information at the moment. We need to decide how much info we need and how to store with as little over head as possible.

Tony – What information would you need from knowing the layer information?

Xin – For example, we can tell whether there is a hit in the first or third layer of the MVTX

Tony – I support that suggestion

Jin – Having a hit in the first layer of the MVTX has a direct impact on our DCA line shape so we do need to know this. We can then split into subsamples based on hits to study the line shapes.

 

Jin – For the DST formats, in general we need to distill the DST down to about 1% of the current size. We need to go from O(100PB) to 1PB so we can have a quick readback, e.g. the SvtxTrack will likely not be there, instead we will use something like a TClonesArray for better compression

 

Xin – Can we have the outer most track projection for calorimeter matching?

Jin – I agree with Xin that the best idea is to also have the outer most layer track state (as well as the beam line state).

 

DST needs, HF-resonances, Vertices

 

Tony – We currently have a list of all the track indices that are associated to each PV but we may not want that. You do need it if you want to loop over all vertices then loop over all tracks from that vertex for an analysis.

Cameron – How do we handle secondary tracks in this scenario?

Tony – This would have to be handled in a resonance container.

 

Xuan – I would like to echo Tony’s suggestions and use a vertex matrix like what was used in PHENIX.

 

DST needs, HF-resonances, Resonances

 

Jin – My guess would be that we will include resonances in the DST but as a separate DST. The reasons are two fold,  we will likely need to rerun the construction for a serious analysis so this avoids needing to delete a container every time on the node tree and will be faster to run over offline

Tony – My worry is the context for resonances, what analysis are we going to do that doesn’t care about the rest of the event?

Cameron – We could check things like momentum resolutions and scale wrt mass

Joe – We can also use it to check the alignment

 

Xin – From experience, second processing passes are challenging as you need to balance time and disk space. It is good to have containers for things like Ks0 reconstruction

 

Joe – Doe we have a PV association in the container?

Xin – I agree that we need this

Cameron – I don’t think KFParticle supports this so we would need to add it to a track-based container

 

DST needs, HF-resonances, Headers

 

Joe – What is an “event number” in this context

Cameron – It would be the trigger frame

Tony – We need to think carefully about this. The only information we will really have about what event a track belongs in is the BC. Tracks can also overlap trigger frames due to edge effects

Cameron – It sounds like this association has to happen at the processing stage

Tony – Yes, I think so

 

DST needs, HF-jets

 

Xuan – For the calorimeters, are we only saving the tower info or is it cluster information as well? I agree with Xin that we need to the projections to the outer tracking layer

Jin – I think the base info is just the towers and the clusters could be added during the production from a module. I don’t know the final decision from the jet TG about how this will be processed.

 

Xin – To come back to my earlier point, I was suggesting adding the outer TPC info to the track info for the calorimeter projections.

Jin – I think the cleanest solution would be to have both the inner and outer most track info to allow analysers to choose which value to use

 

Xuan – Can we also have the projection from the EMCal to the HCal?

Jin – If we have the outer tracking hit then we can do any projection. Are you suggesting a tool to allow users to make any projection they need?

Xuan – Yes

Joe – We do save all the track state information so we could have a module to strip out all or some parts of the track state

 

DST needs, Quarkonia

 

Tony – We can assemble a large electron conversion sample with we can use for calibration and background estimations

Marzia – I think you can kill a lot of the electron conversions by requiring 3 MVTX hits

Tony – Right now the tracking keeps everything (full tracks and TPC only track)

Marzia – With TPC-only, you can’t get the right calorimeter association?

Tony – In triggered mode, you can (Addition from Cameron – I may have missed a chunk of this conversation as I don’t understand what I wrote here, for quarkonia we will use triggered data to get the calorimeter information)

 

From: Cameron Thomas Dean <cameron.dean AT cern.ch>
Date: Tuesday, December 19, 2023 at 10:54 AM
To: sphenix-hf-Jets-l AT lists.bnl.gov <sphenix-hf-jets-l AT lists.bnl.gov>, Huang, Jin via sPHENIX-upsilons-l <Sphenix-upsilons-l AT lists.bnl.gov>, sphenix-physics-l AT lists.bnl.gov <sphenix-physics-l AT lists.bnl.gov>
Subject: Next HF&Q meeting

Dear all,

This is a quick reminder that we will meet tomorrow at 1:30pm ET.

 

 The agenda is as follows:

• Introduction, conveners
• Run 24 upsilon projections¶ - Sasha Lebedev
• Tools required for run 24 analyses and DST format, conveners

The indico link is https://indico.bnl.gov/event/21537/

And the zoom link is https://bnl.zoomgov.com/j/1613938007?pwd=QjYzUnpYQ2xNZWZLSy9qd3Z4dFpzZz09

 

We look forward to talking to everyone tomorrow,
Cameron, Tony, Marzia and Jin

 

From: Cameron Thomas Dean <cameron.dean AT cern.ch>
Date: Tuesday, December 12, 2023 at 8:32 AM
To: lebedev AT iastate.edu <lebedev AT iastate.edu>
Cc: sphenix-hf-Jets-l AT lists.bnl.gov <sphenix-hf-jets-l AT lists.bnl.gov>, Huang, Jin via sPHENIX-upsilons-l <Sphenix-upsilons-l AT lists.bnl.gov>, sphenix-physics-l AT lists.bnl.gov <sphenix-physics-l AT lists.bnl.gov>
Subject: Postponed: Next HF&Q meeting

Dear all,

Apologies for some confusing emails but we’ll postpone the meeting by one week to the 20^th of December. We feel this will allow more time to digest last week’s collaboration meeting and to prepare thoughts for the next run.

Also, we have added Sasha Lebedev to the schedule to discuss upsilon projections for next year as well.

The connection information remains the same.

Cheers,
Cameron, Tony, Marzia and Jin

 

From: Alex Lebedev <lebedev AT iastate.edu>
Date: Monday, December 11, 2023 at 8:11 PM
To: Cameron Thomas Dean <cameron.dean AT cern.ch>
Cc: sphenix-hf-Jets-l AT lists.bnl.gov <sphenix-hf-jets-l AT lists.bnl.gov>, Huang, Jin via sPHENIX-upsilons-l <Sphenix-upsilons-l AT lists.bnl.gov>, sphenix-physics-l AT lists.bnl.gov <sphenix-physics-l AT lists.bnl.gov>
Subject: Re: [Sphenix-upsilons-l] Next HF&Q meeting

Hi Cameron,

I would like to present some slides about our estimates for the run 2024 Upsilon measurement.

Could you add me to the agenda?

Sasha.

 

On Mon, Dec 11, 2023 at 2:00 PM Cameron Thomas Dean via sPHENIX-upsilons-l <sphenix-upsilons-l AT lists.bnl.gov> wrote:

Dear quarkonia and heavy flavor enthusiasts,

We think that last weeks collaboration meeting stimulated lots of good discussion and we would like to keep the momentum for our next meeting. With that in mind, we will have a meeting this Wednesday at 1:30pm ET (13^th of December). Currently the agenda is as follows:

 

• Introduction, conveners
• Tools required for run 24 analyses and DST format, conveners

 

Please let us know if there;s anything that you would like to add to the agenda.

The indico link is https://indico.bnl.gov/event/21537/

And the zoom link is https://bnl.zoomgov.com/j/1613938007?pwd=QjYzUnpYQ2xNZWZLSy9qd3Z4dFpzZz09

 

We look forward to talking to everyone on Wednesday,
Cameron, Tony, Marzia and Jin

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