STAR Flow, Chirality and Vorticity PWG

[Huan Huang <huang AT physics.ucla.edu>]

Hi Han-sheng,

  Unfortunately, the CME background is not due to your dynamical correlation (I call eccentricity correlation) alone. Emission pattern fluctuations are a major source to background as well. Our goal is to suppress background from the combination of these. 

  If you do shape selection based on eccentricity (dynamics in your language) alone, then the most round event shape class corresponds to the most central collisions. In this case, there are few spectator protons and the B field will be minimal, the CME will naturally disappear. That  again defeats the purpose of event shape selection for CME searches.

  Because we want to suppress the combined background sources, so the results will depend on details of the shape selection procedure. In our methodology paper, we discussed several methods. You claimed that you do not know what physics meanings are for the intercepts in our approach. If you read our paper carefully, you may notice that within the AVFD model, the intercept values reproduce what is expected from the CME input in the model. If you change the run parameters of the AVFD model such as you did for viscosity coefficients, can the same optimal ESS approach still work nicely? I do not expect so. Because we want to suppress a combination of background sources, we will need to work with simulations as close to real data as possible. Because we need to suppress the overall background from a combination of sources, I am sure that you can point out that one component of the background is not optimally suppressed versus the other (your argument for resonances versus single particle v2). That is expected and we know that.

  In many occasions, you and Fuqiang insisted that you want q2 and v2 dynamically (eccentricity) correlated alone for shape selection, I explained, at least to Fuqiang a few times, about our concern for the COMBINATION of background sources and why eccentricity selection for most round events get you most central collisions with minimal B field. Our ESS method is intended for CME searches only. I hope that you will think more about our arguments and we can go beyond the repeating of the argument about dynamical (eccentricity) correlation.

  Thanks. Regards,

  Huan

On Fri, Jun 21, 2024 at 7:55 AM Han-sheng Li <li3924 AT purdue.edu> wrote:

Hi Huan,

Thanks for your prompt reply.

q2 and v2 are correlated. If you use an eta gap to calculate q2 and v2, they are correlated through eccentricity. That is good for hydro-related studies. But not enough to remove CME background.

q2 and v2 are correlated more when we use the same eta region of POI, they are correlated through both eccentricity and particle emission pattern. We want to select an event class of azimuthally round events, that will allow us to reduce the CME background.

In ESE, q2 and v2 are correlated via dynamical fluctuations. With varying q2 event classes, average <v2_resonance> varies proportionally to <v2_finalstate>.

In ESS, q2 and v2 are correlated primarily via statistical fluctuations. With varying q2 event classes, it is not guaranteed that <v2_resonance> varies proportionally to <v2_finalstate>.

It is easy to understand if v2_finalstate is the event selection variable in place of q2: plotting v2_resonance vs. v2_finalstate will still have a positive intercept (see PRC95,051901). With the pair q2 selection, it is not clear what intercept v2_resonance vs. v2_finalstate will have. Our model studies suggest the intercept can be positive, zero, and negative.

q2-pair and v2-single are correlated too, and the correlation strength is different from that of q2-single and v2-single.  We found that empirically selecting event classes using q2-pair and calculating delta-gamma from v2-single approaching ZERO will suppress the background in CME search best. Perhaps that is due to the fact that with delta-gamma correlation we deal with more than one particle in the correlation calculation.

Han-Sheng: You seem to be worried about auto-correlation or self-correlation in these quantities. If q2 and v2 are not correlated at all, how do you envision that the event shape method would work?

Yes, q2 and v2 need to be correlated in order for ESE to work. What one needs is the dynamical correlation between the two, not the self-correlation between the two because of their using the same set of particles.

Best regards,

Han-Sheng

---

寄件者: Huan Huang <huang AT physics.ucla.edu>
寄件日期: 2024年6月19日 下午 12:35
收件者: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
副本: Gang Wang <gwang AT physics.ucla.edu>; Han-sheng Li <li3924 AT purdue.edu>
主旨: Re: [Star-fcv-l] 回覆: 回覆: FCV PWG meeting on 19/June/2024 Wed. 9:30 AM EDT

 

---- External Email: Use caution with attachments, links, or sharing data ----

Hello All,

  q2 and v2 are correlated. If you use an eta gap to calculate q2 and v2, they are correlated through eccentricity. That is good for hydro-related studies. But not enough to remove CME background.

  q2 and v2 are correlated more when we use the same eta region of POI, they are correlated through both eccentricity and particle emission pattern. We want to select an event class of azimuthally round events, that will allow us to reduce the CME background.

  q2-pair and v2-single are correlated too, and the correlation strength is different from that of q2-single and v2-single.  We found that empirically selecting event classes using q2-pair and calculating delta-gamma from v2-single approaching ZERO will suppress the background in CME search best. Perhaps that is due to the fact that with delta-gamma correlation we deal with more than one particle in the correlation calculation.

  Han-Sheng: You seem to be worried about auto-correlation or self-correlation in these quantities. If q2 and v2 are not correlated at all, how do you envision that the event shape method would work?

  I am pleased that you look into the ESS and compare with ESE. We need to understand the approaches well. More studies will always help.

  Thanks. Regards,

  Huan

On Wed, Jun 19, 2024 at 3:08 PM Han-sheng Li via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:

Hi Gang,

Thanks for your comments on my slides.

Please find my reply bellow:

Regarding the first bullet on slide 5, if we follow your derivation, shouldn't it be \delta/N^2 instead of \delta/N? 

Total number of quadralets is ~N^4. Total number of triplets is ~N^3, so it does seem to be 1/N scaling. 

Anyway, your derivation is fairly crude and only considers two particles with very similar pT, which is uncommon. 

The derivation isn't meant to be precise. It's to illustrate there's an self-correlation issue within the q2 variable itself which can have an effect on Delta gamma.

The idea of using pair q2 is to eliminate the self-correlation between v2 and q2 on an event-by-event basis. 

q2 and v2 are self-correlated because they are using the same set of POI's. That's why v2 fluctuates over a large range as q2 is varied over a large range, primarily on statistical fluctuations. The statistical fluctuations of v2 and the statistical fluctuations of q2 are correlated with each other.

The so-called self-correlation term (delta/N) you mentioned is not a concern. 

The \delta/N or \delta/N^2 term, even if it exists, doesn't contribute to the self-correlation between v2 and pair q2. 

We think it may bias the Delta gamma as the q2 quantity contains a delta-like contribution. Again, the self-correlation between v2 and pair q2 is because they are calculated using the same set of particles. Their statistical fluctuations are correlated.

It probably even helps us in destroying the self-correlation we want to remove.

Also, note that in our analysis, the spectator plane is used to calculate v2, instead of two-particle correlation v2^2. 

In our model studies, we used the known reaction plane Psi=0 for the calculation of single v2 (cos(2phi)). The v2{2} is only used as a static constant for the normalization of q2^2.

Thus, I'm not sure how relevant your derivation is to our ESS approach. 

As a suggestion for the simulation, it may be useful to try more peripheral events since you are concerned about the larger 'self-correlation' effect there.

The main issue with the ESS approach is that it is not proven to be only sensitive to CME because of the intertwined q2-v2-Delta gamma variables. We don't know what the remaining background contamination exactly is. The self-correlation in the q2 calculation is only one possible issue, probably not even a major issue.

Best regards,

Han-Sheng

---

寄件者: Gang Wang <gwang AT physics.ucla.edu>
寄件日期: 2024年6月19日 上午 07:27
收件者: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
副本: Han-sheng Li <li3924 AT purdue.edu>
主旨: Re: [Star-fcv-l] 回覆: FCV PWG meeting on 19/June/2024 Wed. 9:30 AM EDT

 

---- External Email: Use caution with attachments, links, or sharing data ----

Hi, Han-Sheng

Sorry I missed your presentation, but I reviewed your slides with great interest. 

Regarding the first bullet on slide 5, if we follow your derivation, shouldn't it be \delta/N^2 instead of \delta/N? 

Anyway, your derivation is fairly crude and only considers two particles with very similar pT, which is uncommon. 

The idea of using pair q2 is to eliminate the self-correlation between v2 and q2 on an event-by-event basis. 

The so-called self-correlation term (delta/N) you mentioned is not a concern. 

The \delta/N or \delta/N^2 term, even if it exists, doesn't contribute to the self-correlation between v2 and pair q2. 

It probably even helps us in destroying the self-correlation we want to remove.

Also, note that in our analysis, the spectator plane is used to calculate v2, instead of two-particle correlation v2^2. 

Thus, I'm not sure how relevant your derivation is to our ESS approach. 

As a suggestion for the simulation, it may be useful to try more peripheral events since you are concerned about the larger 'self-correlation' effect there.

Thanks

Gang Wang

Department of Physics and Astronomy

UCLA

On Wed, Jun 19, 2024 at 6:13 AM Han-sheng Li via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:

Dear Convenors,

Please find my slides in the following link:

https://drupal.star.bnl.gov/STAR/blog/hshengli/Model-study-Event-Shape-method

Best regards,

Han-Sheng

---

寄件者: Han-sheng Li <li3924 AT purdue.edu>
寄件日期: 2024年6月18日 下午 01:18
收件者: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
副本: subhash <subhash AT rcf.rhic.bnl.gov>
主旨: 回覆: [Star-fcv-l] FCV PWG meeting on 19/June/2024 Wed. 9:30 AM EDT

 

Dear Convenors,

I would like to present my model study for the event shape method of delta gamma.
Please add me to the agenda.

Best regards,

Han-Sheng

---

寄件者: Star-fcv-l <star-fcv-l-bounces AT lists.bnl.gov> 代表 subhash via Star-fcv-l <star-fcv-l AT lists.bnl.gov>
寄件日期: 2024年6月16日 下午 09:30
收件者: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
副本: subhash <subhash AT rcf.rhic.bnl.gov>
主旨: [Star-fcv-l] FCV PWG meeting on 19/June/2024 Wed. 9:30 AM EDT

 

---- External Email: Use caution with attachments, links, or sharing data ----


Dear All,

We shall have our weekly FCV PWG meeting this Wednesday (19/June/2024)
at 9:30 AM EDT. If you wish to present please let us know. The agenda
will be collected at:
https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Fdrupal.star.bnl.gov%2FSTAR%2Fblog%2Fjjiastar%2Fbulkcorr&data=05%7C02%7Cli3924%40purdue.edu%7C60e1c115c546459bfc5108dc8e9f75d9%7C4130bd397c53419cb1e58758d6d63f21%7C0%7C0%7C638542062766965848%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=eFB37UcDtglKM5Kr%2B7JKZpsn6%2F0EfoNcoO6jhfsvgf4%3D&reserved=0
Please send us your slides by Tuesday. Zoom details are copied below.

Thanks and regards,
Prithwish, Zhenyu and Subhash


ZOOM LINK FOR FCV MEETING:
Join ZoomGov Meeting
https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Fbnl.zoomgov.com%2Fj%2F1612377416%3Fpwd%3DV3kvcnN5ZTRLVEc4U01QWUUycDQ1UT09&data=05%7C02%7Cli3924%40purdue.edu%7C60e1c115c546459bfc5108dc8e9f75d9%7C4130bd397c53419cb1e58758d6d63f21%7C0%7C0%7C638542062766975039%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=kbT6Ub%2B3FG75oJB45hiHlf9qReyT5p9DbPqzJpIj%2FUI%3D&reserved=0

Meeting ID: 161 237 7416
Passcode: 106847

One tap mobile
+16692545252,,1612377416#,,,,*106847# US (San Jose)
+16468287666,,1612377416#,,,,*106847# US (New York)

Dial by your location
         +1 669 254 5252 US (San Jose)
         +1 646 828 7666 US (New York)
         +1 551 285 1373 US
         +1 669 216 1590 US (San Jose)
Meeting ID: 161 237 7416
Passcode: 106847
Find your local number: https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Fbnl.zoomgov.com%2Fu%2FabVqdu5fbU&data=05%7C02%7Cli3924%40purdue.edu%7C60e1c115c546459bfc5108dc8e9f75d9%7C4130bd397c53419cb1e58758d6d63f21%7C0%7C0%7C638542062766981263%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=02Ie0ERry81byV620O4KIhGFnnAW9Wn%2BMAL1Wkd7nQ8%3D&reserved=0

Join by SIP
1612377416 AT sip.zoomgov.com

Join by H.323
161.199.138.10 (US West)
161.199.136.10 (US East)
Meeting ID: 161 237 7416
Passcode: 106847
_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Flists.bnl.gov%2Fmailman%2Flistinfo%2Fstar-fcv-l&data=05%7C02%7Cli3924%40purdue.edu%7C60e1c115c546459bfc5108dc8e9f75d9%7C4130bd397c53419cb1e58758d6d63f21%7C0%7C0%7C638542062766987164%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=hyjdgoEOunwe5VdpTTvVzr0XA%2FCTLcOGQkMwWVOBt0A%3D&reserved=0

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l