STAR HardProbes PWG

[gaohan_yang <gaohan_yang AT m.scnu.edu.cn>]

Hi Yi,

Thank you for your green light. 

The reason we use pp pt shape for isobar efficiency estimation is that we don't obverse the R_{AA} results have significant pT dependence at both RHIC and LHC energies. Meanwhile, we agree we need to spend more efforts to make sure the pt shape won't introduece effieincy bias for future publication.

Best regards,

Gaohan

	gaohan_yang
gaohan_yang AT m.scnu.edu.cn

---- Replied Message ----

From Yi Yang<yiyang0429 AT gmail.com> Date 4/23/2023 08:00 To gaohan_yang<gaohan_yang AT m.scnu.edu.cn> Cc star-hp-l AT lists.bnl.gov<star-hp-l AT lists.bnl.gov> Subject Re: [Star-hp-l] Preliminary figures request on Upsilon production in Isobar collisions.

Hi Gaohan,

Thanks a lot for the updated version. I don't have any further comments and I think your preliminary requests are good to be released. 

However, I am still a bit concerning about the statement "And the p_T distribution for Upsilons, we uesed the shape corresponding in p+p collisions are derived from world-wide data.". I don't think it is good to use the pT shape from the pp collisions (world-wide data) on heavy-ion data, since we don't know if the pT shapes are the same in heavy-ion and pp collisions. We need to be very careful about it... some uncertainties or correction will be needed, but this can be handled during the publication. 

Cheers,

Yi

On Sat, Apr 22, 2023 at 12:02 PM gaohan_yang <gaohan_yang AT m.scnu.edu.cn> wrote:

Dear conveners,

In case you missed my updated, there is a gentle reminder. If you have any comments please feel free let me know.

Best Wishes,

Gaohan

	gaohan_yang
gaohan_yang AT m.scnu.edu.cn

---- Replied Message ----

From	gaohan_yang via Star-hp-l<star-hp-l AT lists.bnl.gov>
Date	4/20/2023 18:17
To	yiyang0429 AT gmail.com<yiyang0429 AT gmail.com>
Cc	star-hp-l AT lists.bnl.gov<star-hp-l AT lists.bnl.gov>
Subject	Re: [Star-hp-l] Preliminary figures request on Upsilon production in Isobar collisions.

Hello Yi,

Thanks very much for your nice comments, please find the updated(slides 7):
https://drupal.star.bnl.gov/STAR/blog/ghyang/Preliminary—Upsilon-analysis-isobar-collisions-1

Best,

Gaohan

	gaohan_yang
gaohan_yang AT m.scnu.edu.cn

---- Replied Message ----

From	Yi Yang<yiyang0429 AT gmail.com>
Date	4/20/2023 15:38
To	gaohan_yang<gaohan_yang AT m.scnu.edu.cn>
Cc	barbara.trzeciak AT gmail.com<barbara.trzeciak AT gmail.com> , star-hp-l AT lists.bnl.gov<star-hp-l AT lists.bnl.gov>
Subject	Re: [Star-hp-l] Preliminary figures request on Upsilon production in Isobar collisions.

Hi Gaohan,

Thanks a lot for updated slides and I have two minor follow-up: 

 - p7: as s a function --> as a function 

 - Can you add the n_sigma_e in different eta bins and the fits in the backup slides? 

Cheers,

Yi

On Thu, Apr 20, 2023 at 1:43 PM gaohan_yang <gaohan_yang AT m.scnu.edu.cn> wrote:

Hi, Yi

Thank you for your valuable comments. I have implemented them and uploaded new version on the Drupal page which can be found in this link:

https://drupal.star.bnl.gov/STAR/blog/ghyang/Preliminary%E2%80%94Upsilon-analysis-isobar-collisions-1

Please see my detailed responses inline:

| -p6: for track quality cut: are you applying 3.5 GeV/c cut on both electrons' pT, or just the triggered electron (leading one)? 

— yeah, we applying 3.5 GeV/c cut on both electrons' p_T.

|-p7: please define f(eta) 

— done (see it in s7).

| -p8: How do you get the efficiencies for different pT bins? Are you assuming a certain pT distribution for Upsilons?

— The Usplon efficiencies for different pT bins were calculate by：number of reconstructed  𝛶s(p_T) with analyzed cuts/ number of embedded 𝛶s(p_T) 

And the p_T distribution for Upsilons, we uesed the shape corresponding in p+p collisions are derived from world-wide data.

| -p11: Just for my own education, is there any specific reason using the exponential form to fit the combinatorial background? 

          Any other constraints on the fits? For example N(2S)? 

— We used two function to fit combinatorial background, one is (default) and the others is  . The parameters for each of them are getting from the unbinned maximum-likelihood simultaneous fit to unlike-sign and like-sign. And we don't add such constraints on the fits.

 |-p15: I am a bit surprised that the systematics from track quality cuts are up to ~15%, do you know why? I remember it should be around 10% or lower? 

* I remember we discussed in one of the PWG meetings about the acceptance dependency on polarization. Any follow-up on this? 

— Yes, you are right. The systematics uncertainties from track quality I showd before is around 10%. But there is a STAR new guidence suggested that everyone add 5% uncertainty on single track sefficiency that why out systematics uncertainties from track quality cuts are up to ~15%.

Best,

Gaohan 

On Wed, Apr 19, 2023 at 3:54 PM Barbara Trzeciak via Star-hp-l <star-hp-l AT lists.bnl.gov> wrote:

Hi Gaohan,

thanks. The preliminaries look good to me.

Cheers,

Barbara

On Tue, Apr 18, 2023 at 7:12 PM gaohan_yang <gaohan_yang AT m.scnu.edu.cn> wrote:

Hi, Barbara

I appreciate your useful common.  I have implemented them and uploaded new version on the Drupal page which can be found in this link: 

https://drupal.star.bnl.gov/STAR/blog/ghyang/Preliminary%E2%80%94Upsilon-analysis-isobar-collisions-1

Please see my detailed responses inline:

|- s16,17: Isobar\sqrt{s_NN} -> Isobar \sqrt{s_NN} (missing space)

— done (see it in s17, 18).

|- s17: Since the global uncertainties here are just from N_coll I suggest: global uncertainty -> N_coll uncertainty. Also, They should be the same for Upsilon(1S) and Upsilon(2S) so it's enough if you plot only one (e.g. in gray or some other color).

I would also add a legend explaining what are different boxes (AuAu and pp uncertainties) - it's not a paper figure where you have it explained in the caption. 

— done (see it in s18). To be consist with Au+Au and other plots, we just add pp uncertainties legend.

Could you please prepare a version of this plot where you compare to the Au+Au results. 

— Sure, I will prepare this plot in next few days which should not affect the preplinary request approval。

|- s18: I think you can keep this plot from 0 on the y-axis for the preliminary result - I wanted to see internally how it looks like. 

 I would suggest moving "STAR Preliminary" and "Isobar ..." inside the plot to be consistent with your other plots. 

The black dashed line is not explained in the legend. 

— done (see it in s16).

|- s19: Since you have the same y and pT coverage for both Isobar and Au+Au, you don't need to repeat this info twice in the legend, you can separate it out. 

— done (see it in s19).

Could you please compare also result for Upsilon(2S) state

— I understand where you're coming from, considering the large uncertainties of Y(2S) in isobar collisions, the comparision plot won't make significant physics impact. We prefer not to put them together.

| have you tried to e.g. require that the Upsilon(2S) yield is larger than Upsilon(3S) and see what happens there ? Also, for this pT bin you have wider inv. mass binning due to less statistics, but have you checked the fits with narrower binning 

—We don't require that the Upsilon(2S) yield is larger than Upsilon(3S). And we used unbined fitting, just binning width is drawn. It actually has no effect on fit.

Best,

Gaohan

	gaohan_yang
gaohan_yang AT m.scnu.edu.cn

---- Replied Message ----

From	Barbara Trzeciak<barbara.trzeciak AT gmail.com>
Date	4/18/2023 19:29
To	gaohan_yang<gaohan_yang AT m.scnu.edu.cn>
Cc	star-hp-l AT lists.bnl.gov<star-hp-l AT lists.bnl.gov> , syang AT m.scnu.edu.cn<syang AT m.scnu.edu.cn>
Subject	Re: [Star-hp-l] Preliminary figures request on Upsilon production in Isobar collisions.

Hi Gaohan,

I have also questions regarding your inv. mass fitting - slides 20-21.

You don't have much Upsilon(3S) signal, which is understandable, but there are some bins where the behaviour is less expected - in particular the last pT bin on slide 21 where you get negative Upsilon(2S).

Have you tried to e.g. require that the Upsilon(2S) yield is larger than Upsilon(3S) and see what happens there ? Also, for this pT bin you have wider inv. mass binning due to less statistics, but have you checked the fits with narrower binning ?

Cheers,

Barbara

On Tue, Apr 18, 2023 at 9:34 AM Barbara Trzeciak <barbara.trzeciak AT gmail.com> wrote:

Hi Gaohan,

thank you for the clarifications and updated slides. 

They look good to me, we can discuss fine tuning of some points, like the trigger sys. unc. towards the publication. 

Please see below my minor comments to your preliminary plots.

- s16,17: Isobar\sqrt{s_NN} -> Isobar \sqrt{s_NN} (missing space)

- s17: Since the global uncertainties here are just from N_coll I suggest: global uncertainty -> N_coll uncertainty. Also, They should be the same for Upsilon(1S) and Upsilon(2S) so it's enough if you plot only one (e.g. in gray or some other color).

I would also add a legend explaining what are different boxes (AuAu and pp uncertainties) - it's not a paper figure where you have it explained in the caption. 

Could you please prepare a version of this plot where you compare to the Au+Au results. 

- s18: I think you can keep this plot from 0 on the y-axis for the preliminary result - I wanted to see internally how it looks like. 

 I would suggest moving "STAR Preliminary" and "Isobar ..." inside the plot to be consistent with your other plots. 

The black dashed line is not explained in the legend. 

- s19: Since you have the same y and pT coverage for both Isobar and Au+Au, you don't need to repeat this info twice in the legend, you can separate it out. 

Could you please compare also result for Upsilon(2S) state

Cheers,

Barbara

On Mon, Apr 17, 2023 at 8:06 PM gaohan_yang <gaohan_yang AT m.scnu.edu.cn> wrote:

Hello Babara,

Thank you for your valuable comments. I have implemented all of them and uploaded new version on the Drupal page which can be found in this link:

https://drupal.star.bnl.gov/STAR/blog/ghyang/Preliminary%E2%80%94Upsilon-analysis-isobar-collisions-1

 Please see my detailed responses inline.

|- s11: I guess you use template fitting for all three Upsilon states, I see that you get -4 Upsilon(3S) - could you please include a plot where you go to negative count values, so that we can see what happens for Upsilon(3S). Also, please add to the backup fits for all the centrality and pT bins.

— done (see it in slides 21 and 22).

|- s12: how do you decide on the variation values for the a parameter of the smearing ?

— We used the same method employed in the upsilon measurements in the AuAu collisions. The variation values is justified by observing the minimum chi2 distribution. The same value was used in our preliminary Jpsi results in isobar collisions.

|- s12: could you please add on the plots values of the uncertainties for each of the varied case. 

— done (see them in slides 12, here I used Upsilon(1S) and put the Upsilon(2S) in the backup ).

|- s13: do you vary the cuts separately or simultaneously, and is the final sys. unc. max or RMS value ?

— We varied the cuts separately and take the final RMS value as the systematic value. We made this point clear now in the updated slides.

|- s14: is the final sys. unc. max or RMS value of the variations ?

— We took the RMS as the final sys. unc. We made this point clear in the updated slides.

|- s15: how do you estimate the trigger selection criteria uncertainty, and couldn't it differ for the two Upsilon states ?

— Instead of dsmAdc (online), adc0 (offline) is used as the selection criteria for trigger electron in this analysis. The dsmAdc is a truncated value of adc0 with pedestal subtraction. From simulation, there exists that events are with adc0>300 but dsmAdc ≤ 18, which could be caused by the pedestal subtraction. We conservatively assigned the difference as the systematic uncertainty arising from trigger electron selection. We agree that the excited Upsilon is likely to more easily to trigger BEMC, but the difference should be subdominant.

|- s16: Let's leave the N_coll uncertainties like this for now, to be consistent with Au+Au.

Add from where is the p+p reference. What global uncertainties include ?

— OK. We used the same pp reference employed in our new published upsilon paper. For R_{AA} vs. Npart, the global uncertainties include the total Y(1S+2S+3S) uncertainty (stat. & sys., vast majority global uncertainty) from combining STAR and PHENIX measurements, the uncertainties of yield ratio among different Upsilon states from the world wide data. For the R_{AA} vs pT, besides above two sources, we also include the uncertainties from the p_T shape parameterization.

|- s16: could you please add a version of this plot compared to the Au+Au results. 

— done, see it in sllides19.

|- s17: What global uncertainties include ? They look quite large, I would expect here that they are from N_coll unc., and p+p uncertainties are included in the pT dependent Upsilon unc. 

— You are absolutely correct. We accidentally used the AuAu N_{coll} uncertainty, now we fixed this issue see it in slides17.

|- s18: Add legends for data points and fit lines. Add Upsilon decay channel. 

— done, see it in sllides18.

|- s16-17: Add physics conclusions to your preliminary physics results. 

— done, see it in sllides20.

By the way, I have also seen your reply to my presentation for ATHIC 2023. I am revising the slides and will get back to you as soon as possible. Thank you very much for your patience and very useful advice.

Best,

Gaohan

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