sphenix-physics-l AT lists.bnl.gov
Subject: sPHENIX discussion of physics
List archive
Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu
- From: "Perepelitsa, Dennis" <dvp AT bnl.gov>
- To: Cameron Thomas Dean <cameron.dean AT cern.ch>
- Cc: "sphenix-physics-l AT lists.bnl.gov" <sphenix-physics-l AT lists.bnl.gov>, "Huang, Jin via sPHENIX-HF-jets-l" <sphenix-hf-jets-l AT lists.bnl.gov>
- Subject: Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu
- Date: Sat, 28 Jan 2023 17:56:34 +0000
Hi Cameron,
Thanks for this detailed write-up which contains our current snapshot of HF reconstruction in Au+Au collisions and a roadmap to the eventual first measurements, and thanks also for your patience for my returning these a day late.
Clearly the document represents a tremendous amount of effort!
I have some questions and suggestions below. Since this is an internally-facing document, I tried to aim towards increasing the clarity and adding more detailed information that our collaborators could benefit from.
I was able to read through the main body for this first circulation, but I think it would be useful to have a presentation in an upcoming HF/Upsilon TG meeting about the material in Appendix B.
Dennis
—
Abstract
“roughly 20 minutes” -> wouldn’t 25 minutes be a much closer approximation?
Section 1
Line 4: “calculation” -> “reconstruction” ?
There is something funny with Eq. 1 and thus Eq. 2:
If this is meant to be for fully min-bias A+A collisions and L is the AA luminosity, L^AA, then the way to covert to the equivalent nucleon-nucleon luminosity (which is what multiplies sigma_ccbar) is L^NN = A^2 L^AA, i.e. the geometric factor is A^2 and not Ncoll.
However, if these are meant to be centrality-selected A+A collisions, the equivalent “nucleonic luminosity” is Nevt^AA * TAA, where TAA = Ncoll / sigma_NN. I suppose you could convert the L^AA to an Nevt using the total Au+Au cross-section and then a centrality fraction.
So either of the above could work, but I think as written, Eqs. 1 and 2 are not quite right.
Lines 17-18: is it worth noting something already here about the partial correlation of the acceptance and tracking efficiencies between the two tracks? i.e. it is an approximate but not perfect factorization into the product of single-track efficiencies.
Line 42: “are detailed”
Section 2
What was the vertex fixed position or distribution? Does that impact the extracted efficiencies (particularly the acceptance)?
I would be more technical on the Pythia setup - were these SoftQCD:nonDiffractive processes (?) but then filtered on the presence of a ccbar at any point in the event record’s history?
Line 65: “Almost” -> “Approximately” ? “Almost” sounds like we fell short of some goal.
Line 74: Can you describe the hit duplication in a bit more detail? Is it exactly the same hit pattern, or it is the hit pattern from a different (unrelated) event?
Line 81: “below” is unambiguous - you don’t have to add “in the vertical direction”
Section 3
Since we note the evolution of KFParticle, it would be good to specify that DecayFinder and HFTrackEfficiency are internal sPHENIX packages - maybe even with citations that have links to the git areas for the modules?
Line 117: apostrophe after “particles” ? And similarly, “mothers” on L152
Line 118: “geometrical” -> “geometric” ?
Line 150, 153, and possibly others: maybe it is just me, but I found it funny to list “truth momenta” and then “eta” as a distinct additional piece of information. The pseudorapidity is an aspect of the four-momentum.
Line 154: “tracking efficiency for heavy flavor particles” — we don’t track the HF particles themselves, maybe better to write “tracking efficiency for the decay products of … “ ?
Line 156-160: just to clarify, is this triggering capability used in the generation of the simulation samples? Or we are just noting a feature of the module?
Line 159-160: I didn’t quite understand this. If this is only done when all tracks are reconstructed, how does that allow users to build a decay “without any selections”? Requiring the reconstruction is a rather restrictive set of selections…
Section 4
Line 164: I think it would be useful to say a bit about how the PV is built.
Line 172: Can you remind us what DIRA Is an acronym or abbreviation for?
Line 172-173: “quality of the separation”, “quality of the track IP” are a bit unambiguous. How can a separation have a “quality”? Maybe better to write “quality of the estimate of the separation” - ?
Footnote 4: Can we write the matching limit out - is it, i.e., | (px^truth - px^reco) / px^truth| < 0.05 and similarly for the other two directions? I think this is what L184 is saying, but better to be explicit when first introduced.
Figure 1: I think it’s very useful to have this Figure, and here are some suggestions which may hopefully make it even clearer: (1) the arrow for pT(D0) is red and points up towards the variable name, while the analogous other arrows are black and point down - better to standardize?, (2) the double-sided red arrow which indicates the DCA is drawn at a funny spot which does not visually look like the place where the trajectories come the closest together - anything we can do here? (3) I would suggest to label the gray arrow, which I think the momentum of the mother D
Line 190: “hypothetical K-pi+ pair”? Since this is really the background.
Line 190 and others: The reader can probably guess, but it would be better to just include units (GeV) on mass ranges.
Line 192: “is it” -> “are that it”
Line 198 and others: for the six-digit numbers, it would be useful to add either a comma or a LaTeX separation \, to break them up into groups of three digits
Line 202: “, and the associated”
Line 204: is that the signal efficiency, or the signal-to-background ?
Figure 2, right panel: This sideband extends down to 1.95 GeV, whereas I thought in the text we specified 2.0 GeV.
Figure 3 caption: “dashed” -> “hatched” ?
Figure 5: the text describes one of these curves, but not the other two?
Table 2: It comes off as a bit funny to see quantities of length given in very small fractions of a cm — one would think micrometers are more relevant. Or is this because cm is the default unit in KFParticle (or some other configurable module) ?
Table 2: for the first four cuts, you specify whether it is a min or max requirement, but then not for the next four. One can make intelligent guesses (<chi^2, >pT, etc.) but would be nice to just specify.
Section 5
Line 221: do I understand correctly that under the swapped mass hypothesis, the mean of the distribution is still expected to be the same as for the correct hypothesis? Is that really right? Might be useful to make that explicit.
Naively, for every correct-mass pairing, there is an attempted swapped-mass pair as well, and thus k_cor-ID = 1/2 exactly. I guess the reason it is > 50% is because the swapped-mass distribution is broader than the mass window, while the correct-mass distribution is fully contained. Might be useful to just spell that out.
L258: typo “hypothesis”
Section 6
Line 271: I was a bit confused by the choice of |eta| < 1.6? (And I believe that is |eta|, not eta, right?) Is it useful to have such a large region in the truth definition where sPHENIX cannot possibly measure particles? The acceptance comes out looking artificially quite small.
Line 280: again, should this be |eta| ?
Line 284: Can you give the reader more context about the associated photon? Is this from >2 body decays where there is a pi0 or photon, or this is a Brem photon off one of the charged hadrons?
Figure 8, left panel: the bins towards the upper right oscillate a lot in their central values. Do we understand those as statistical fluctuations (i.e. compatible with the stated uncertainties)? Is it better perhaps to rebin? Otherwise, what information is conveyed by the very jumpy bin-to-bin efficiency values?
Figure 8, right: would it be better to add a right-axis for the blue distribution?
Figure 8 caption: ends with a trailing “..”
Table 4: would it be clearer to specify “fails pT but not eta”, “fails eta but not pT” ? Do we also need the entries for the associated pi0 - if there are zero counts, one could just note this in the text…
Table 5 caption: “all tracks”
Figure 9: I still don’t quite understand the point of having the definition go out to |eta| < 1.6. Of course there will be many D0 pairs not in the acceptance where those pairs tracks in the region eta = 1-1.6, because sPHENIX does not intend to measure there! It seems tautological. The more interesting cases seem to be when the min track pT is large and/or the max eta is small — what is happening there? But the plot is dominated by these “artificial” cases of inefficiency.
Tables 7 and 8: funny sig figs issues - e.g. 1867.45 +/- 2.1
L324: missing period after 2 GeV
L331: “One is the”
L332: typo “efficiency”
L336: “appear to be”
L344: why does the K0s have a softer pT spectrum than the D0?
Abstract
“roughly 20 minutes” -> wouldn’t 25 minutes be a much closer approximation?
Section 1
Line 4: “calculation” -> “reconstruction” ?
There is something funny with Eq. 1 and thus Eq. 2:
If this is meant to be for fully min-bias A+A collisions and L is the AA luminosity, L^AA, then the way to covert to the equivalent nucleon-nucleon luminosity (which is what multiplies sigma_ccbar) is L^NN = A^2 L^AA, i.e. the geometric factor is A^2 and not Ncoll.
However, if these are meant to be centrality-selected A+A collisions, the equivalent “nucleonic luminosity” is Nevt^AA * TAA, where TAA = Ncoll / sigma_NN. I suppose you could convert the L^AA to an Nevt using the total Au+Au cross-section and then a centrality fraction.
So either of the above could work, but I think as written, Eqs. 1 and 2 are not quite right.
Lines 17-18: is it worth noting something already here about the partial correlation of the acceptance and tracking efficiencies between the two tracks? i.e. it is an approximate but not perfect factorization into the product of single-track efficiencies.
Line 42: “are detailed”
Section 2
What was the vertex fixed position or distribution? Does that impact the extracted efficiencies (particularly the acceptance)?
I would be more technical on the Pythia setup - were these SoftQCD:nonDiffractive processes (?) but then filtered on the presence of a ccbar at any point in the event record’s history?
Line 65: “Almost” -> “Approximately” ? “Almost” sounds like we fell short of some goal.
Line 74: Can you describe the hit duplication in a bit more detail? Is it exactly the same hit pattern, or it is the hit pattern from a different (unrelated) event?
Line 81: “below” is unambiguous - you don’t have to add “in the vertical direction”
Section 3
Since we note the evolution of KFParticle, it would be good to specify that DecayFinder and HFTrackEfficiency are internal sPHENIX packages - maybe even with citations that have links to the git areas for the modules?
Line 117: apostrophe after “particles” ? And similarly, “mothers” on L152
Line 118: “geometrical” -> “geometric” ?
Line 150, 153, and possibly others: maybe it is just me, but I found it funny to list “truth momenta” and then “eta” as a distinct additional piece of information. The pseudorapidity is an aspect of the four-momentum.
Line 154: “tracking efficiency for heavy flavor particles” — we don’t track the HF particles themselves, maybe better to write “tracking efficiency for the decay products of … “ ?
Line 156-160: just to clarify, is this triggering capability used in the generation of the simulation samples? Or we are just noting a feature of the module?
Line 159-160: I didn’t quite understand this. If this is only done when all tracks are reconstructed, how does that allow users to build a decay “without any selections”? Requiring the reconstruction is a rather restrictive set of selections…
Section 4
Line 164: I think it would be useful to say a bit about how the PV is built.
Line 172: Can you remind us what DIRA Is an acronym or abbreviation for?
Line 172-173: “quality of the separation”, “quality of the track IP” are a bit unambiguous. How can a separation have a “quality”? Maybe better to write “quality of the estimate of the separation” - ?
Footnote 4: Can we write the matching limit out - is it, i.e., | (px^truth - px^reco) / px^truth| < 0.05 and similarly for the other two directions? I think this is what L184 is saying, but better to be explicit when first introduced.
Figure 1: I think it’s very useful to have this Figure, and here are some suggestions which may hopefully make it even clearer: (1) the arrow for pT(D0) is red and points up towards the variable name, while the analogous other arrows are black and point down - better to standardize?, (2) the double-sided red arrow which indicates the DCA is drawn at a funny spot which does not visually look like the place where the trajectories come the closest together - anything we can do here? (3) I would suggest to label the gray arrow, which I think the momentum of the mother D
Line 190: “hypothetical K-pi+ pair”? Since this is really the background.
Line 190 and others: The reader can probably guess, but it would be better to just include units (GeV) on mass ranges.
Line 192: “is it” -> “are that it”
Line 198 and others: for the six-digit numbers, it would be useful to add either a comma or a LaTeX separation \, to break them up into groups of three digits
Line 202: “, and the associated”
Line 204: is that the signal efficiency, or the signal-to-background ?
Figure 2, right panel: This sideband extends down to 1.95 GeV, whereas I thought in the text we specified 2.0 GeV.
Figure 3 caption: “dashed” -> “hatched” ?
Figure 5: the text describes one of these curves, but not the other two?
Table 2: It comes off as a bit funny to see quantities of length given in very small fractions of a cm — one would think micrometers are more relevant. Or is this because cm is the default unit in KFParticle (or some other configurable module) ?
Table 2: for the first four cuts, you specify whether it is a min or max requirement, but then not for the next four. One can make intelligent guesses (<chi^2, >pT, etc.) but would be nice to just specify.
Section 5
Line 221: do I understand correctly that under the swapped mass hypothesis, the mean of the distribution is still expected to be the same as for the correct hypothesis? Is that really right? Might be useful to make that explicit.
Naively, for every correct-mass pairing, there is an attempted swapped-mass pair as well, and thus k_cor-ID = 1/2 exactly. I guess the reason it is > 50% is because the swapped-mass distribution is broader than the mass window, while the correct-mass distribution is fully contained. Might be useful to just spell that out.
L258: typo “hypothesis”
Section 6
Line 271: I was a bit confused by the choice of |eta| < 1.6? (And I believe that is |eta|, not eta, right?) Is it useful to have such a large region in the truth definition where sPHENIX cannot possibly measure particles? The acceptance comes out looking artificially quite small.
Line 280: again, should this be |eta| ?
Line 284: Can you give the reader more context about the associated photon? Is this from >2 body decays where there is a pi0 or photon, or this is a Brem photon off one of the charged hadrons?
Figure 8, left panel: the bins towards the upper right oscillate a lot in their central values. Do we understand those as statistical fluctuations (i.e. compatible with the stated uncertainties)? Is it better perhaps to rebin? Otherwise, what information is conveyed by the very jumpy bin-to-bin efficiency values?
Figure 8, right: would it be better to add a right-axis for the blue distribution?
Figure 8 caption: ends with a trailing “..”
Table 4: would it be clearer to specify “fails pT but not eta”, “fails eta but not pT” ? Do we also need the entries for the associated pi0 - if there are zero counts, one could just note this in the text…
Table 5 caption: “all tracks”
Figure 9: I still don’t quite understand the point of having the definition go out to |eta| < 1.6. Of course there will be many D0 pairs not in the acceptance where those pairs tracks in the region eta = 1-1.6, because sPHENIX does not intend to measure there! It seems tautological. The more interesting cases seem to be when the min track pT is large and/or the max eta is small — what is happening there? But the plot is dominated by these “artificial” cases of inefficiency.
Tables 7 and 8: funny sig figs issues - e.g. 1867.45 +/- 2.1
L324: missing period after 2 GeV
L331: “One is the”
L332: typo “efficiency”
L336: “appear to be”
L344: why does the K0s have a softer pT spectrum than the D0?
On Jan 13, 2023, at 6:25 PM, Cameron Thomas Dean <cameron.dean AT cern.ch> wrote:
<D0_in_AuAu_sPHENIX_C_Dean.pdf>Hi Dennis,
Oops, I should’ve remembered that. I’ve attached a version with line numbers.
Cheers,Cameron
On Jan 13, 2023, at 6:10 PM, Perepelitsa, Dennis via sPHENIX-HF-jets-l <sphenix-hf-jets-l AT lists.bnl.gov> wrote:
Hi Cameron,_______________________________________________
Thanks for sending this around. Can I ask you to re-circulate to the lists but with a line-numbered version of the draft?
That would help with giving comments.
Dennis
On Jan 13, 2023, at 4:32 PM, Cameron Thomas Dean <cameron.dean AT cern.ch> wrote:
<D0_in_AuAu_sPHENIX_C_Dean.pdf>_______________________________________________Dear colleagues,
Following my presentation in today’s heavy flavor and quarkonia meeting, I’ve attached the analysis note I wrote on reconstructing D0’s during the commissioning period.
I will circulate this in the sPHENIX physics mailing lists for 2 weeks to collect any comments before circulating it in the general mailing list.
Cheers,
Cameron
sPHENIX-physics-l mailing list
sPHENIX-physics-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-physics-l
Dennis V. Perepelitsa
Associate Professor, Physics Department
University of Colorado Boulder
on sabbatical at Brookhaven National Laboratory
sPHENIX-HF-jets-l mailing list
sPHENIX-HF-jets-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-hf-jets-l
Dennis V. Perepelitsa
Associate Professor, Physics Department
University of Colorado Boulder
on sabbatical at Brookhaven National Laboratory
Associate Professor, Physics Department
University of Colorado Boulder
on sabbatical at Brookhaven National Laboratory
-
[Sphenix-physics-l] sPHENIX ana note: D0 in AuAu,
Cameron Thomas Dean, 01/13/2023
-
Re: [Sphenix-physics-l] sPHENIX ana note: D0 in AuAu,
Perepelitsa, Dennis, 01/13/2023
-
Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu,
Cameron Thomas Dean, 01/13/2023
- Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu, Xin Dong, 01/20/2023
- Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu, Perepelitsa, Dennis, 01/28/2023
-
Re: [Sphenix-physics-l] [Sphenix-hf-jets-l] sPHENIX ana note: D0 in AuAu,
Cameron Thomas Dean, 01/13/2023
-
Re: [Sphenix-physics-l] sPHENIX ana note: D0 in AuAu,
Perepelitsa, Dennis, 01/13/2023
Archive powered by MHonArc 2.6.24.