STAR HardProbes PWG

["Mooney, Isaac" <isaac.mooney AT yale.edu>]

Hi Neil,

I already sent an email to Sooraj and Barbara requesting the GPC formation, so you should be hearing from Sooraj soon. See below for a couple of small responses to your responses.

Thanks,

Isaac

On Dec 6, 2024, at 13:50, Diptanil Roy <roydiptanil AT gmail.com> wrote:

Hi Isaac, thank you for your vote of confidence. Please let me know what you need from my end to form the GPC. Once there is the go-ahead from the PWGC, and I get some space to commit my code, I can do that.

Please find the responses to your comments below. Since the comments are minor and related to the previous comments, I have updated the AN link in place.

Okay good, but can you make this clear in the text of the analysis note?

Yes, reorganized the text to make this clear. (L193)

 But now that I’m looking at the plot again — you have only a few bins in your raw distributions near z ~ 0 - 1, but have 7 bins in your corrected z results. Is there a reason you can’t bin the raw data a bit finer around this region, since it looks like you have the statistics to do so? I know that migration means uncorrected z = 0 - 1 is not the same as corrected z = 0 - 1, but it still seems that there is some room for increased granularity. Have you looked at it?

Important caveat: What I am showing are the projections of 2D distributions - I found a somewhat optimal binning in 2D such that very few bins have < 10 entries. I could of course increase the granularity a bit, and during the early stages of the analysis, I experimented a bunch and found the unfolding to be generally stable with respect to the binning within reason.

Let me just make sure I understand your comment about the momentum resolution: if you lose e.g. the kaon, you know the momentum of the particle-level kaon, you know the momentum resolution for kaons, so you randomly sample from this distribution for a given momentum to get the detector-level momentum and assign this value?

Yes, that's absolutely correct.

I’m not sure I understood this last line. You do e.g. R->Miss(pTgen,w)? But the weight is based on the comparison of detector-level pT to data. So accessing the weight distribution using a generator-level pT is not appropriate, right?

I don't use the generator level pT, I still use the detector level pT for the reconstructed jet whose axis is outside our pseudorapidity acceptance. Of course, at the very edge of that acceptance, the jets might start being less than conical, but such misses are only 5% of the total sample, so I don't think this has an outsized effect on the unfolding. 

>> Okay, I’m not sure I’m understanding exactly the procedure here. It would be great if you could prepare a slide to include in your first presentation to the GPC that presents the step-by-step unfolding process in fine detail, and we can have a discussion then. 

 Okay, but just for my curiosity do you have plots of the (raw) 2D correlations between the three observables? E.g. DeltaR and z? 

Nope, unfortunately, never set that up since I was always unfolding in conjunction with the jet pt. I am adding this to the TODO list during the GPC process.

> Okay, thanks. By the way, in the new version the typo is still there.

Little confused by this. Do you mean I should change from 2.8 cm to 2 cm in the text? If so, I would prefer to change it after I implement it to avoid confusion.

>> I just mean flip “>” to “<“ in l. 439.

Ah, I see. I think the text is a bit unclear, with the word “this” being ambiguous. I would recommend just “Our final source of systematic uncertainty is given by the second term on the right hand side…” 

Done.

 Ah that makes sense. Thanks for clarifying. But why isn’t the vertex correction uncertainty explicitly mentioned in the D0-meson reconstruction related uncertainties section? I see it is mentioned in the D0 analysis note. Anyway, it would be good to add some explanation somewhere in the analysis note of what this uncertainty/legend means (similar to what you told me here).

Yep, added a bit to the D0-meson reconstruction related uncertainties section. (L420)

On Wed, Dec 4, 2024 at 12:48 PM Mooney, Isaac <isaac.mooney AT yale.edu> wrote:

Hi Neil,

Thanks for your responses. For the most part, I’m satisfied (although I had some responses inline below). I agree with Nihar that we can request the GPC formation for this mature analysis now.

Thanks,

Isaac

On Dec 1, 2024, at 08:14, Diptanil Roy <roydiptanil AT gmail.com> wrote:

Dear Issac and Qian,

Thank you for your detailed comments on the AN. Please find my replies inline below:

P.S. Seems like Drupal is down (including all STAR webpages), so I am attaching the latest version of the AN to the next email. I will update the webpage when drupal is available again.

Issac's comments:

==============================================================================================================================

79. How was 600 MeV chosen for the D0 daughters?

This is the cut that was chosen for the D0 spectra and the D0 v2 analyses. I am not entirely sure if this cut was picked after an analysis of the S/B ratio or in an ad-hoc way, but the effect of this high pt cut for the D0 daughter candidates (as opposed to say 0.2 or 0.3 GeV/c) is to suppress the combinatorial background in the low pt region for the KPi pairs.

83. What is the nhitsdEdx cut used?

None, again this is for consistency with the D0 spectra and the D0 v2 analyses. I checked both the ANs along with the codes available with the paper submissions in CVS. Again, the nHitsFit > 20 is a pretty strict cut, so a separate nHitsdEdx cut might not be essential.

139. Sorry, just to make sure I understand, do you mean "all the jet constituents" except the D0 candidate, or really "all" jet constituents? I'm not sure why we wouldn't assign the proper mass in a situation where we're fairly sure of that particle's identity.

First reason is that using the mass of the D0 instead of pi0 doesn't affect any of the quantities we are measuring. The comparison for PYTHIA jets for jet pT is below.

<image.png>

Second reason is that in the hadronic correction, we always remove the energy contribution of the tracks by assuming pion mass. To keep things consistent, I chose to use the pi0 mass. 

In my opinion, the only quantity that should be affected by this should be the jet mass. If someone looks at that in the future, it might be prudent to treat the mass in a better way.

Fig. 1.5. I'm a little surprised that the weights are quite this negative in the background regions. But I guess this is due to a negative covariance and may not be a problem as long as the sum of sWeights in a given control variable bin is non-negative? Are there any such bins in the analysis where the sum is negative?

There aren't any negative bins for D0 pT, however, if we plot a 2D distribution of D0 Jet pT and Z and fill the entries with these sWeights, there could be some bins where the count is negative. I have chosen the binnings such that there are as few of these bins as possible, however when they exist, the count is set to zero, and the error is set to the actual error of the bin.

168. Is the reason there is any discrepancy at all entirely due to sPlot? Or are there any other differences (e.g. selection criteria, binning, failed jobs, different random seed, etc.)?

Most of the above. The selection criteria is similar, however the fitting method in sPlot is an unbinned maximum likelihood fit (as opposed to binned fits for the old analyses), some of the files used in 2018 are no longer available in the distributed disks etc.

 189. When you apply the hadronic correction, have the K and pi already been removed from consideration so that their track pTs are not subtracted from the tower ETs, or do you include them for this correction and remove them later? I would assume the latter except you've already mentioned in the text that they are removed.

Yes, the latter is true. I account for the K pi energy that is deposited on the towers and perform hadronic correction on them in the same way as I do for all other tracks.

> Okay good, but can you make this clear in the text of the analysis note?

Fig. 1.10. I can understand that because sequential recombination algorithms in general don't have to have an exact cone shape at the jet radius (although anti-kT is fairly circular), you could technically get D0s beyond DeltaR = R_jet. But DeltaR = 0.6 would be pretty surprising to me. Or do the entries in the bin stop at just a hair past 0.4, and the bin is just wider than necessary? 

It's the latter, the bin is quite large at the end, and is removed after unfolding anyway.

Figs. 1.10 - 1.12. I'm a bit surprised the z_jet^uncorrected plots look identical between Figs. 1.11 and 1.10. Looking at the 40-80% distributions as an example, I'm not able to spot the difference. I would have thought in the tails, either positive or negative, there should be some difference, despite the much lower counts of D0jets with D0 pT > 5 GeV, since the D0 pT being larger should contribute to a larger z. I see from Fig. 1.12 that actually the tails drop off much more dramatically for the larger D0 pT so I guess my intuition was wrong. Should I think of this as being caused by the fact that jets with a large pT D0 are much less likely to be background jets, which decreases counts both for negative and large-positive z? 

Yeah, I needed to do a double-take on this as well. First thing is we should probably not read too much into trends in uncorrected z plots. There's a lot of pT smearing, so the trends are not trivial to begin with. But even if we just look at the jet pt distribution for D0 pT [1-5], [5-10] and [1-10] together (plot below), we can see the 3 orders of magnitude difference between D0 pT [1-5] and [5-10] GeV (1st panel). The 2nd panel is just the ratio ([1-5] + [5-10])/[1-10] spectra, and is 1 as expected.  

<Screenshot 2024-12-01 at 5.21.08 AM.png>

Corollary 1: is the jet pT fixed for the D0 z plot, or is it integrated over jet pT? [Whatever the answer, it would be good to specify somewhere in the analysis note, similarly for the DeltaR as well].

These are all the D0 jets, no cuts on the jet pT. When I unfold, I unfold over the whole jet pt range. For example, for D0 pT [1-10] GeV, the jet pT can physically take values above 1 GeV. So, I unfold the whole spectra from pT,Jet > 1 GeV, and then quote the final spectrum from 5 - 20 GeV only. I added a note to clarify this in the AN (L206 - 208).

 

Corollary 2: is there any requirement on number of jet constituents? Is it possible to have a jet that is just the D0 if it passes all other requirements? I think we talked about this during a PWG meeting and you mentioned that these one-particle jets were included (so there's a separate category included in the 0.9 - 1 z bin which are identically 1.0) [I just read the paper and see this is the case]. Have you checked the effect that it has on the results (pT, z, DeltaR) to disallow these jets?

There is no requirement on the number of jet constituents. You recall correctly that I included an overflow bin (z in [1-1.1]) in an iteration, and did the unfolding accordingly and the results of the unfolding were unchanged. I have not removed jets like these completely, because I wasn't sure how to account for the bias introduced in the unfolding due to removing the single particle jets. Also, another analysis of fragmentation function for J/psi from CMS here also kept the single particle jets in the mix by setting z = 1 to z = 0.999 with a converged unfolding, so we used the same in this analysis.

> Ah okay, taking these three answers together helps me understand a bit better the behavior of the z_jet^uncorrect plots. Thanks.

Fig. 1.14. I think I understand why the distribution in raw z is bimodal for the 40-80% centrality selection for the previous plots, but do you have a good intuition for why it isn't continuous in the 10-40% case for R = 0.2? And where are the bins just below 0 for the 40-80%?

I am not sure if I follow the statement about discontinuity for 10-40 % alone. If you mean the non-differentiable curve around 0, that's true for the z distributions in all centralities. One of the bins is z in [-0.5, 0] and the other is z in [0, 0.5]. These have contributions from different jets. These have contributions from different jets. 

1. z in [-0.5, 0] comes from jets where too much energy had been removed from the jets using the area based background subtraction resulting in corrected pT,Jet around the left tails of the pT,Jet distributions.
2. Accordingly, z in [0., 0.5] comes from jets where not a lot of energy had been removed from the jets using the area based background subtraction resulting in corrected pT,Jet around the right tails of the pT,Jet distributions.

For the bin below 0 for the 40-80 %, the count was 0 for those bins from the sPlot method.

> I’m also not sure if I follow my statement :). I think it was getting a bit late at the time. But now that I’m looking at the plot again — you have only a few bins in your raw distributions near z ~ 0 - 1, but have 7 bins in your corrected z results. Is there a reason you can’t bin the raw data a bit finer around this region, since it looks like you have the statistics to do so? I know that migration means uncorrected z = 0 - 1 is not the same as corrected z = 0 - 1, but it still seems that there is some room for increased granularity. Have you looked at it?

275. When you say "already corrected for the D0 reconstruction efficiency at this stage" do you mean there is a step between the previous and this one of applying D0 efficiency?

I mean, in the measured data, the uncorrected spectra are "already corrected for the D0 reconstruction efficiency ...". So, when we unfold, this is not an effect we should take into consideration. Therefore, we make sure that every D0 generated in PYTHIA at the particle level is available for analysis in the detector level. Of course, we can not do that by requiring this condition on the reconstructed event for GEANT (i.e. the kaon and pion from D0 decay must be matched to two distinct reconstructed tracks in the detector level, this introduces unnecessary bias in the sample which is not well-explained, and also would also be a waste of precious compute). Therefore, we take the momentum resolution of the kaon pion tracks from simulation and always reconstruct the D0 daughters in the detector level, irrespective of the presence of one or both of the daughter tracks (which are discarded). 

> I understand (I think). To me it just made sense to have bullets discussing what you do to get a response; and also have a bullet describing what you do to the data before it’s fed into the unfolding. This would include correcting for D0 reconstruction efficiency. But I see that above you specifically say “to get a response matrix”, not “to unfold”, so it’s probably fine how it is.

> Let me just make sure I understand your comment about the momentum resolution: if you lose e.g. the kaon, you know the momentum of the particle-level kaon, you know the momentum resolution for kaons, so you randomly sample from this distribution for a given momentum to get the detector-level momentum and assign this value?

278. I'm a little confused -- you say the D0 is reconstructed at detector-level but the daughters are not? And then you use momentum resolution to reconstruct it? Can you clarify what you mean here?

Hopefully, the last answer cleared this one up. The way we ensure the D0 is always present at the reconstructed level is by smearing the particle level kaon and pion daughters from PYTHIA with the momentum resolution and using those smeared tracks to reconstruct the D0. Simply "triggering" on D0 events which eventually have a reconstructed D0 would not be ideal for reasons mentioned above.

Fig. 1.20 - 1.22. It would be great if the y-axis of the ratio plots could be zoomed in a bit so any potential shapes could be observed. Something like 0.5 - 1.5 should work in all cases given how excellent the agreement is.

Done. Replaced the plots with the recommended axis limits in the AN now. (Pg 31 - 33)

374. Are fakes also weighted by this factor? What, if anything, is done for misses? 

The way I have set up the unfolding is to include all jets reconstructed with a D0 inside in the response matrix, so there are no fakes due to the jet pT. This is the reason for such wide acceptance ranges in the uncorrected jet pt plots (Figs 1.10-1.14). Fakes can also happen if a D0 jet is inside the pseudorapidity acceptance range in the detector level but is outside in the particle level. In the sample we have, such cases are almost negligible (mostly 0 for all cases), hence these are ignored. 

Misses can happen because a particle level D0 jet with the pseudorapidity acceptance range is reconstructed as a detector level jet outside the pseudorapidity acceptance range. In such cases, the weight factors (dependent only on pT and z) are used to fill the response matrix.

> I’m not sure I understood this last line. You do e.g. R->Miss(pTgen,w)? But the weight is based on the comparison of detector-level pT to data. So accessing the weight distribution using a generator-level pT is not appropriate, right?

390. Can this statement be quantified somehow? Can you show the correlations between the three observables: pT, z, and DeltaR for example? Or show the 3D unfolding result even though statistics are poor? Or some other way.

So, the 3D unfolding results do not converge at all for this analysis, ergo it's quite difficult to do that. However, we do still unfold with pT vs DeltaR distribution after reweighting by the data-driven fragmentation function. The variation due to this is about 2-3 % in most cases (see Fig. 1.36, 1.45) for pT,D < 5 GeV/c and about 10 % (see Fig. 1.51) for pT,D > 5 GeV/c. This is probably the best way to show the effect of the prior variation on the radial profile.

> Okay, but just for my curiosity do you have plots of the (raw) 2D correlations between the three observables? E.g. DeltaR and z? 

415. Hijing -> Pythia? Or is Hijing used somehow?

No, this is Hijing used by the D0 spectra analysis only. They used a data-driven simulation method and also a full sample of HIJING, and this uncertainty is to deal with the differences between the two. See Fig. 81 here along with the associated text. 

431. "DCA < 3 cm", yes? 0.2 cm seems like a fairly minimal variation. Do statistics really significantly suffer for e.g. a 2 cm cut as a variation?

Yes, 0.2 cm is small, and I had checked with 2 cm as well for QM2023. The statistics suffer with a cut of 2 cm for the peripheral cases more than most, however the results were consistent. I can redo this in the background with the latest updates to the code to show the effect of reducing DCA to 2 cm.

> Okay, thanks. By the way, in the new version the typo is still there.

462. You say that only the uncertainty on R1 is quoted but also say the uncertainty is given by the second term on the RHS of eq. 1.7. How are these both true?

 

The statistical uncertainty % is quoted from R1 only. The 2nd term on the RHS of eq. 1.7 is the uncertainty due to a variation of the prior. This is systematic in nature and hence is quoted as such.

> Ah, I see. I think the text is a bit unclear, with the word “this” being ambiguous. I would recommend just “Our final source of systematic uncertainty is given by the second term on the right hand side…” 

Fig. 1.32.
This probably demonstrates that I'm not as familiar as I should be with the previous D0 analysis, but can you explain why you have uncertainties related both to D0 reconstruction efficiency with and without vertex correction?
When you say "peripheral (right)", you actually mean "bottom" in this case, right? Since the plot has wrapped to the next row. Just want to make sure I'm not mixing up midcentral and peripheral.

 This is just to differentiate the uncertainty from the topological cuts vs that from the vertex correction part. These are both derived from the D0 spectra analysis. And yes, peripheral (right) is actually "bottom". I have updated the captions now.

> Ah that makes sense. Thanks for clarifying. But why isn’t the vertex correction uncertainty explicitly mentioned in the D0-meson reconstruction related uncertainties section? I see it is mentioned in the D0 analysis note. Anyway, it would be good to add some explanation somewhere in the analysis note of what this uncertainty/legend means (similar to what you told me here).

Fig. 1.65. Shouldn't the full jet pT spectrum be harder than the charged jet spectrum? It looks like it is until the last one or two bins in all cases, which is a bit odd to me.

There is an additional cut on the D0 pT of 1-10 GeV/c which means the total number of charged jets in the pT,Jet range (3-20) GeV is more or less consistent with the total number of full jets in the pT,Jet range (5-20) GeV. Even with that, the full jets spectrum is harder in all cases other than the last bin for central events.

> Hmm, it still seems like the distributions collapse to each other above 13 or so GeV, which is beyond the upper pT limit of the D0 so it can’t just be that the jet is only composed of those two daughter tracks meaning the full jet is the same as the charged jet. I need to think about this a bit more. 

Qian's comments

==============================================================================================================================

1. It don’t show a discussion of sWeight’s error bar and how the error
bars propagated in your final results?

Since the sWeights are just entered into the histograms as weights, they are propagated as Poisson errors i.e error of a bin = sqrt(sum of all weights in that bin). In the trivial case where all the weights are 1, this is the sqrt(n) error. This is handled by ROOT histograms directly when sumw2 is called before filling histograms, hence I didn't add a separate discussion.

 

2. It is not clear to me how did you avoid the double count for the
towers in Jet reconstruction part. How did you carry out the hadronic
correction?

Each track is matched to a tower (or to a nearest tower) by the track-tower matching algorithm in the event maker. I calculate the energy contribution of all the matched tracks on to a specific tower by assuming the tracks are all pions and subtracting the energy from the tower's recorded energy. 

I have now included this description in the AN (L191).

3. Jet Pt with background correction will have some negative value. Did
you also included this part in your z and delta R calcuation?

Yes, z is calculated with the corrected jet pt, that's why z takes unphysical negative values. delta R is not directly dependent on jet pt, however, there can be an implicit dependence. Hence, 2D unfolding is used to correct (jet pt, z) and (jet pt, delta R).

4. For the data-driven unfolding method, do you have a plot of proof of
convergence for this method?

The trivial closures still hold for the data-driven unfolding method. Since they are repetitive, I did not include them in the AN. Beyond that, it is just a choice of prior we are making.

<Screenshot 2024-12-01 at 7.43.30 AM.png>

Thank you again for the detailed questions.

On Wed, Nov 27, 2024 at 9:01 PM tc88qy <tc88qy AT rcf.rhic.bnl.gov> wrote:

Hi Neil and PAs,

   Nice work. Since the D0 meson reconstruction is identical to STAR
published paper. I assumed this part should be ok.
I have few comments on your analysis note, please find below:
1. It don’t show a discussion of sWeight’s error bar and how the error
bars propagated in your final results?
2. It is not clear to me how did you avoid the double count for the
towers in Jet reconstruction part. How did you carry out the hadronic
correction?
3. Jet Pt with background correction will have some negative value. Did
you also included this part in your z and delta R calcuation?
4. For the data-driven unfolding method, do you have a plot of proof of
convergence for this method?

Qian Yang

On 2024-11-28 06:05, Mooney, Isaac wrote:
> Hi Neil and PAs,
>
> For now I’m focusing my comments on the analysis note (see below),
> since I will be the PWG rep for this analysis and can give detailed
> comments on the paper after GPC formation. Hopefully this saves a bit
> of time getting the ball rolling due to Neil’s time constraints. I
> did do a quick read-through of the short paper as well just to make
> sure that it’s acceptable for GPC formation. I had some comments,
> but nothing preventing it moving to the next step, so I’ll hold off
> for now until I do a more detailed read-through at that time. We’ll
> see if Qian or Nihar have any major comments, but mine on the analysis
> note can be addressed in parallel with GPC formation which I think the
> analysis is ready for.
>
> Thanks, and congratulations on advancing this high-quality analysis to
> this stage,
> Isaac
>
> 79. How was 600 MeV chosen for the D0 daughters?
>
> 83. What is the nhitsdEdx cut used?
>
> 139. Sorry, just to make sure I understand, do you mean "all the jet
> constituents" except the D0 candidate, or really "all" jet
> constituents? I'm not sure why we wouldn't assign the proper mass in a
> situation where we're fairly sure of that particle's identity.
>
> Fig. 1.5. I'm a little surprised that the weights are quite this
> negative in the background regions. But I guess this is due to a
> negative covariance and may not be a problem as long as the sum of
> sWeights in a given control variable bin is non-negative? Are there
> any such bins in the analysis where the sum is negative?
>
> 168. Is the reason there is any discrepancy at all entirely due to
> sPlot? Or are there any other differences (e.g. selection criteria,
> binning, failed jobs, different random seed, etc.)?
>
> 189. When you apply the hadronic correction, have the K and pi already
> been removed from consideration so that their track pTs are not
> subtracted from the tower ETs, or do you include them for this
> correction and remove them later? I would assume the latter except
> you've already mentioned in the text that they are removed.
>
> Fig. 1.10. I can understand that because sequential recombination
> algorithms in general don't have to have an exact cone shape at the
> jet radius (although anti-kT is fairly circular), you could
> technically get D0s beyond DeltaR = R_jet. But DeltaR = 0.6 would be
> pretty surprising to me. Or do the entries in the bin stop at just a
> hair past 0.4, and the bin is just wider than necessary?
>
> Figs. 1.10 - 1.12. I'm a bit surprised the z_jet^uncorrected plots
> look identical between Figs. 1.11 and 1.10. Looking at the 40-80%
> distributions as an example, I'm not able to spot the difference. I
> would have thought in the tails, either positive or negative, there
> should be some difference, despite the much lower counts of D0jets
> with D0 pT > 5 GeV, since the D0 pT being larger should contribute to
> a larger z. I see from Fig. 1.12 that actually the tails drop off much
> more dramatically for the larger D0 pT so I guess my intuition was
> wrong. Should I think of this as being caused by the fact that jets
> with a large pT D0 are much less likely to be background jets, which
> decreases counts both for negative and large-positive z?
> Corollary 1: is the jet pT fixed for the D0 z plot, or is it
> integrated over jet pT? [Whatever the answer, it would be good to
> specify somewhere in the analysis note, similarly for the DeltaR as
> well].
> Corollary 2: is there any requirement on number of jet constituents?
> Is it possible to have a jet that is just the D0 if it passes all
> other requirements? I think we talked about this during a PWG meeting
> and you mentioned that these one-particle jets were included (so
> there's a separate category included in the 0.9 - 1 z bin which are
> identically 1.0) [I just read the paper and see this is the case].
> Have you checked the effect that it has on the results (pT, z, DeltaR)
> to disallow these jets?
>
> Fig. 1.14. I think I understand why the distribution in raw z is
> bimodal for the 40-80% centrality selection for the previous plots,
> but do you have a good intuition for why it isn't continuous in the
> 10-40% case for R = 0.2? And where are the bins just below 0 for the
> 40-80%?
>
> 275. When you say "already corrected for the D0 reconstruction
> efficiency at this stage" do you mean there is a step between the
> previous and this one of applying D0 efficiency?
>
> 278. I'm a little confused -- you say the D0 is reconstructed at
> detector-level but the daughters are not? And then you use momentum
> resolution to reconstruct it? Can you clarify what you mean here?
>
> Fig. 1.20 - 1.22. It would be great if the y-axis of the ratio plots
> could be zoomed in a bit so any potential shapes could be observed.
> Something like 0.5 - 1.5 should work in all cases given how excellent
> the agreement is.
>
> 374. Are fakes also weighted by this factor? What, if anything, is
> done for misses?
>
> 390. Can this statement be quantified somehow? Can you show the
> correlations between the three observables: pT, z, and DeltaR for
> example? Or show the 3D unfolding result even though statistics are
> poor? Or some other way.
>
> 415. Hijing -> Pythia? Or is Hijing used somehow?
>
> 431. "DCA < 3 cm", yes? 0.2 cm seems like a fairly minimal variation.
> Do statistics really significantly suffer for e.g. a 2 cm cut as a
> variation?
>
> 462. You say that only the uncertainty on R1 is quoted but also say
> the uncertainty is given by the second term on the RHS of eq. 1.7. How
> are these both true?
>
> Fig. 1.32.
> This probably demonstrates that I'm not as familiar as I should be
> with the previous D0 analysis, but can you explain why you have
> uncertainties related both to D0 reconstruction efficiency with and
> without vertex correction?
> When you say "peripheral (right)", you actually mean "bottom" in this
> case, right? Since the plot has wrapped to the next row. Just want to
> make sure I'm not mixing up midcentral and peripheral.
>
> Fig. 1.65. Shouldn't the full jet pT spectrum be harder than the
> charged jet spectrum? It looks like it is until the last one or two
> bins in all cases, which is a bit odd to me.
>
>> On Nov 9, 2024, at 14:49, Diptanil Roy <roydiptanil AT gmail.com>
>> wrote:
>>
>> Dear conveners and HP-pwg,
>>
>> We now have a proposed version of the STAR D0 Meson Tagged Jets
>> paper drafts available. The links for the paper drafts, analysis
>> notes, and the paper webpage are below. We request to form a GPC to
>> get this work over the line.
>>
>> Please send your comments and feedback.
>>
>> Paper Drafts
>>
>> Short Paper: Link [1]
>> Long Paper: Link [2]
>>
>> Analysis Note Link [3]
>>
>> Paper Webpage Link [4]
>>
>> PWGC Preview Link [5]
>>
>> Thank you to everyone who helped with this analysis since the
>> beginning.
>>
>> --
>>
>> ~ Neil on behalf of the PAs (Neil, Matt, Sevil, Joern)
>
>
>
> Links:
> ------
> [1] https://drupal.star.bnl.gov/STAR/system/files/PRL_v1.pdf
> [2] https://drupal.star.bnl.gov/STAR/system/files/PRC_v1.pdf
> [3] https://drupal.star.bnl.gov/STAR/system/files/AnalysisNote_6.pdf
> [4]
> https://drupal.star.bnl.gov/STAR/blog/droy1/D0-Meson-Tagged-Jets-Au-Au-collisions-200-GeV
> [5]
> https://drupal.star.bnl.gov/STAR/system/files/D0Jets_Diptanil_PWGCPreview.pdf

--

~ Neil

--

~ Neil