STAR Flow, Chirality and Vorticity PWG

[ShinIchi Esumi <esumi.shinichi.gn AT u.tsukuba.ac.jp>]

Dear Xiaoyu
Thank you very much for the replacement of the figures. 

About the ADC cut, since you are ignoring amplitude information in the geant 
simulation just to count the number of particle per tile in the simulation, 
while 
you are effectively excluding pedestal noise and small ADC part below 
nMip<0.3 
etc in your experimental fitting method. So I just wanted to make sure you 
are not 
counting particle, if the ADC is quite small in your simulation, this is to 
make it 
as same as the experimental treatment. 

Since there is no reason why not to believe the ADC distribution in the 
simulation, 
therefore I was thinking this would be a good test ground for your nMip 
fitting to 
see how well you can count the number of particle by using the ADC 
distribution, 
which you might have already done with the fast Epd simulator, but this 
provides 
the more realistic ADC distribution coming from the dE/dx and photon 
statistics 
with the true geometry including BG hits in the geant simulation, where we 
could 
even include electronics noise to make reasonable pedestal peak, if we like 
to, 
where we know the truth and measured ADC. 

However your earlier test with UrQMD and fast simulator might include already 
the most of effects, though. Do I understand you correctly for your fast 
simulator 
with UrQMD, you overlay the extracted “measured” single MIP ADC distributions 
according to the number of tracks in a tile and then apply your nMip fitting 
to 
decompose the average number of track from the overlaid ADC distribution 
using 
the same single shape, which should give the same answer in principle by 
definition? This is what you are showing in p9 for your DNP talk? 
Best regards, ShinIchi

> On Oct 23, 2020, at 7:46, Liu, Xiaoyu <liu.6566 AT buckeyemail.osu.edu> wrote:
> 
> Hello Shinichi, 
> 
> Thanks for your suggestions. I updated my DNP slides with the new UrQMD 
> plots (P9: 16 variable eta bins instead of 10 even eta bins) and the new 
> version can be found at the same link: 
> https://drupal.star.bnl.gov/STAR/presentations/dnp-2020/vn-measurement-auau-sqrtsnn-27-gev-event-plane-detector-star-1
> 
> 
> For the v1 focus meeting, here is my slides: 
> https://drupal.star.bnl.gov/STAR/system/files/v1_meeting_102020_0.pdf
> If you are asking about the comparison between dN/deta of the simulated EPD 
> hits (blue points on P5) and the dN/deta in the EPD acceptance extracted 
> from the nMIP distribution with the real data, then I don't have it but I 
> can make one later.
> 
> I didn't use any ADC cut and I am not sure if it is needed. Even the ADC is 
> very small, the signal is still created by some particle hitting the EPD 
> tile therefore it should be counted. The small value is just due to the 
> Landau fluctuation. 
> 
> I am not sure if I understand your last suggestion. Do you mean for each 
> EPD tile, I can measure the number of primary particles (generated 
> particles) hitting the tile and the number of "simulated particles" (EPD 
> hits after going through the STAR material simulation) hitting the tile? 
> What do you want to learn from it?
> 
> Thanks,
> Xiaoyu
> From: Star-fcv-l <star-fcv-l-bounces AT lists.bnl.gov> on behalf of ShinIchi 
> Esumi via Star-fcv-l <star-fcv-l AT lists.bnl.gov>
> Sent: Wednesday, October 21, 2020 10:34 PM
> To: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
> Subject: Re: [Star-fcv-l] STAR presentation by Xiaoyu Liu for DNP 2020 
> submitted for review
>  
> Dear Xiaoyu
> Thank you, that makes sense. This would mean we should not plot the results 
> with too small eta bins even for the input values to be compared with 
> output values. 
> On the other hand, in the real event we do have quite a large z-vertex 
> variation, 
> so eta average of each tile in each event are different, so one could still 
> plot a little 
> finer bins in eta, if we like to, so we need to think about the optimal 
> size of the 
> eta bin. I’m fine with your nice talk for DNP. 
> Best regards, ShinIchi
> 
> PS : So the v1 vs eta in UrQMD and our data are quite different, which 
> could be 
> reasonable, because of the poor description of spectators. Do you also have 
> a 
> comparison of dN/deta between them? 
> 
> PPS : Sorry to ask you additional questions on your another presentation in 
> v1 
> focus meeting yesterday, where you were just binary counting hits in each 
> tile of 
> EPD to get measured v1 in MC, which is indeed reasonable, but did you 
> exclude 
> hits below the value of 0.5*ADC for one mip, that need to be applied after 
> integrating all hits ADC values per tile in each HI event? 
> 
> Since you do have 2 ADC distributions for each tile and for both one mip 
> (one 
> particle event) and multi-mips (overlaid many single events) including 
> landau 
> and photon statistical fluctuation in addition to the real increase of the 
> number 
> of particles between the two, therefore you could also apply your nMip 
> fitting 
> method to see how well you can reconstruct the true “increase of” average 
> number of particle per tile in the MC, which is a kind of good test ground 
> of 
> your approach and would also gives us the systematic error as well. You 
> might 
> need to have some constrained fitting of these two simultaneously, where 
> they 
> would have some common parameters in the function or do the fitting of the 
> multi-mip data with a convoluted "data" function from the single-mip 
> distribution. 
> 
>> On Oct 22, 2020, at 1:44, Liu, Xiaoyu <liu.6566 AT buckeyemail.osu.edu> wrote:
>> 
>> Hello Shinichi, 
>> 
>> Attached are the UrQMD + EPD fast simulator study with 16 variable eta 
>> bins (corresponding to 16 EPD rings). The black points and red points are 
>> almost on top of each other. So the discrepancy at large |eta| on my slide 
>> 5 is indeed due to the fact that the innermost ring covers two eta bins. 
>> 
>> Please let me know if you have any other suggestions or comments.
>> 
>> Thanks,
>> Xiaoyu   
>> From: Liu, Xiaoyu <liu.6566 AT buckeyemail.osu.edu>
>> Sent: Sunday, October 18, 2020 8:07 PM
>> To: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>; 
>> ShinIchi Esumi <esumi.shinichi.gn AT u.tsukuba.ac.jp>
>> Subject: Re: [Star-fcv-l] STAR presentation by Xiaoyu Liu for DNP 2020 
>> submitted for review
>>  
>> Hello Shinichi, 
>> 
>> Thank you for your comments and sorry I didn't show the results at recent 
>> BulkCorr meetings. As you mentioned, for this talk, I just want to focus 
>> on introducing the new method and the v1, v2 plots are indeed unnecessary 
>> for this purpose so I deleted them. 
>> 
>> About the UrQMD study, I used the full Au+Au UrQMD events, not the 
>> particle gun if it is what you are asking. And they are just UrQMD events 
>> w/o going through any simulation for the STAR materials. 
>> 
>> I believe the discrepancy in large |𝜂| is due to the fact that the 
>> inner-most ring covers two largest 𝜂 bins therefore EPD cannot tell 
>> exactly which 𝜂 bin the hit belongs to. So v1_EPDhit (4.8<|𝜂 |<5.1) = 
>> v1_EPDhit (4.5<|𝜂 |<4.8) and they are all between v1_track (4.8<|𝜂 |<5.1) 
>> and v1_track (4.5<|𝜂 |<4.8). As for the bin 4.2<|𝜂 |<4.5, part of it falls 
>> on ring 1 and part of it falls on ring 2 so the v1_EPDhit(4.2<|𝜂 |<4.5) 
>> will be pulled either towards the bin on the left or the bin on the right 
>> therefore v1_EPDhit(4.2<|𝜂 |<4.5) is smaller than v1_track (4.2<|𝜂 |<4.5) 
>> which is the peak of v1. In other words, if I plot the same plot with the 
>> x-axis bins being the 16 EPD rings instead of 10 even 𝜂 bins in [2.1,5.1], 
>> the red points and the black points are expected to be on top of each 
>> other. I can make the plot and show you later.
>> 
>> Sorry again for not discussing the results at BulkCorr earlier. Hope it 
>> will be okay if I only show the UrQMD results in this talk. I can present 
>> the UrQMD study in this week's FCV meeting if needed. 
>> 
>> I uploaded the second version under the same link: 
>> https://drupal.star.bnl.gov/STAR/presentations/dnp-2020/vn-measurement-auau-sqrtsnn-27-gev-event-plane-detector-star-1
>>    
>> 
>> Please let me know if you have any other comments or suggestions.
>> 
>> Thanks,
>> Xiaoyu
>> From: Liu, Xiaoyu <liu.6566 AT buckeyemail.osu.edu>
>> Sent: Sunday, October 18, 2020 7:32 PM
>> To: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>; 
>> ShinIchi Esumi <esumi.shinichi.gn AT u.tsukuba.ac.jp>
>> Subject: Re: [Star-fcv-l] STAR presentation by Xiaoyu Liu for DNP 2020 
>> submitted for review
>>  
>> Hello Jiangyong, 
>> 
>> Thank you for your comments, I implemented most of them except:
>> 
>> P3: I did use a random point on the tile to determine the \eta of a hit 
>> EPD tile.
>> 
>> P5: I changed the "the number of 1-, 2-, k-...MIP events per collision" to 
>> "the fraction....". The reason why I didn't use "probability" is that it 
>> is only the probability when we fit the dN/dnMIP of one tile, in which 
>> case the integral of the dN/dnMIP distribution over nMIP is 1. However, if 
>> we look at the dN/dnMIP of several tiles then the integral of the 
>> distribution will be larger than 1 and the sum of the 1, 2, 3, 4-MIP 
>> weights might be larger than 1 too and the word "probability" can be 
>> misleading. Then I realized the word "number" can be equally misleading 
>> since people would expect the number of events to be integers. Therefore, 
>> I chose "fraction" and hopefully it is less confusing. 
>> 
>> P9: The data points are not mirrored around y=0, they do look very similar 
>> though. 
>> 
>> Besides, I decided to delete the v1 before the resolution correction with 
>> the real data as Shinichi suggested since I want to focus on the new 
>> method and these plots are not essential for this purpose.
>> 
>> I uploaded the second version to the same link: 
>> https://drupal.star.bnl.gov/STAR/presentations/dnp-2020/vn-measurement-auau-sqrtsnn-27-gev-event-plane-detector-star-1
>>    
>> 
>> Please let me know if you have any other comments or suggestions.
>> 
>> Thanks,
>> Xiaoyu 
>> 
>> 
>> From: Star-fcv-l <star-fcv-l-bounces AT lists.bnl.gov> on behalf of ShinIchi 
>> Esumi via Star-fcv-l <star-fcv-l AT lists.bnl.gov>
>> Sent: Saturday, October 17, 2020 11:54 PM
>> To: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
>> Subject: Re: [Star-fcv-l] STAR presentation by Xiaoyu Liu for DNP 2020 
>> submitted for review
>>  
>> Dear Xiaoyu
>> 
>> I could sign off your nice talk, but you have not discussed this in the 
>> PWG yet, so it would be nice that you could have shown us before. 
>> 
>> P7-8 : The statistical errors are invisible, but do you know why the 
>> difference 
>> between track(pure sim) and hit(including geant) is suddenly increasing at
>> |eta|>4. Are they a strong detector effect from a large dE/dx from 
>> spectator?
>> You use the true R.P. angle for both cases, is this geant simulation done 
>> taking a full AuAu event of UrQMD or taking a single track from the UrQMD 
>> event one by one? 
>> 
>> P9-10 : Do you like your data plots to be approved as preliminary even 
>> with 
>> the uncorrected data? (for E.P. resolution and BG and detector effects) 
>> If you insist to show these, I would at least label them more clearly that 
>> this 
>> is uncorrected for many possible other effects, not just for R.P. 
>> resolution) 
>> If you like to explain your analysis methods in detail, you would not need 
>> to 
>> use the experimental data, but can be done more with UrQMD data. It seems 
>> we do see a strong spectator contributions in both v1 and v2 (and 
>> most-likely 
>> in dN/deta distribution as well) in data, but you can not really defend 
>> what 
>> this is just with uncorrected data, so I wonder what we can say with these 
>> plots. 
>> 
>> Best regards, ShinIchi
>> 
>>> On Oct 18, 2020, at 10:55, Jiangyong Jia via Star-fcv-l 
>>> <star-fcv-l AT lists.bnl.gov> wrote:
>>> 
>>> Hi, xiaoyu,
>>> 
>>> Here is my comments for sign off
>>> cheers,
>>> Jiangyong
>>> 
>>> P1: please add "for the STAR Collaboration" and only keep your name
>>> 
>>> P2: For outline consider avoiding technical terms such as "nMIP" and 
>>> StEpdFastSim in
>>> 
>>> P4: I would swap the order the two sub-bullet of the nMIP
>>> 
>>> P5: Fit to extract "the probability of 1-2..k-MIP events" in bin j
>>> 
>>> P6: please indicate the AuAu 27 GeV on this figure and also other figures 
>>> where appropriate. 
>>> resolution correction --> correct for the event plane resolution
>>> 
>>> P7: StEpdFastSim--> EPD fast simulation
>>> same format as in the STAR offline data.
>>> 
>>> a random point on the tile--> do you mean the center of the tile?
>>> 
>>> P8: ring one--> the inner-most ring
>>>  to correct for the measurement-->to account for residual detector effects
>>> 
>>> The plots have been symmetrized around y=0, if so please indicate that
>>> 
>>> 
>>> P10: please add some statement on this slide
>>> 
>>> Plots on P8-P10 should have proper x- and y-axis labels
>>> 
>>> P11: we developed a new method
>>> 
>>> 
>>> 
>>> On 10/15/20 11:53 AM, webmaster--- via Star-fcv-l wrote:
>>>> Dear star-fcv-l AT lists.bnl.gov
>>>>  members,
>>>> 
>>>> Xiaoyu Liu (
>>>> liu.6566 AT osu.edu
>>>> ) has submitted a material for a review, please  
>>>> have a look:
>>>> 
>>>> https://drupal.star.bnl.gov/STAR/node/52381
>>>> 
>>>> 
>>>> ---
>>>> If you have any problems with the review process, please contact  
>>>> 
>>>> webmaster AT www.star.bnl.gov
>>>> 
>>>> _______________________________________________
>>>> 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
>> 
>> <Picture1.pdf>