http://cms.cern.ch/cds/JME-13-003
Stefano C.:
Congratulations to the authors: the paper is well written, easy-to-read and understadable even if the topic is far from being simple. Some minor comments below
Type B
- I had sometimes the feeling that this paper has been written by many hands. This is understandable, given the size of the topic, but sometimes it results in repetitions and suboptimal connections between the paragraphs. In addition, in some paragraph there seem to be much more details with respect to others, and it is not always clear the reason. Example will be highlighted in type A comments
- All the plots have “CMS Preliminary 2012”
- In figure 21 the results on the parallel and perpendicular resolution functions are presented. In our understanding, they are expected to agree within the uncertainty among the different channels in which they are measured, namely Z+jets and photon+jets, but clearly they do not, especially for the parallel component. Do you have any hint why? Moreover, the specific algorithms for PU mitigation seem to overperform the PF Et much more in the the Z+jets channels with respect to the photon+jets: why this behaviour? Is it a hint of the correlation between the samples used in the validation and the ones used in the BDT training?
Type A
- Abstract: as a general comment, it is a bit too long and detailed and it may be largely improved in readability
1) there are 8 “missing transverse momentum (energy)”. I know it's hard to find a synonymous for that, but maybe not all are necessary. For instance, in line 9 of the abstract: “missing transverse momentum resolution” → “(its) resolution”
2) “Using these advanced …”: drop this sentence, since the message is already clear from the previous one
L52: X0 is not defined
L145: DeltaR is not defined and will be defined later in another paragraph (see also type A comment about this)
Figure 2: it's not clear why such a plot is important in this context. In addition, why the Data to MC ratio is not drawn here?
L163: eq 1 is a definition, not a condition! You may want to refer to the same cut applied for the muons and introduced in line n.156
L171-176: it would be clearer to first define the “working points” and then use them in the text.
L174: “quality requirements” is a bit generic (see also type A comment)
L180-181: “the standard jet identification criteria” is again too generic and “standard” has a meaning only within the collaboration. Either explain further, or use a reference where those criteria are defined, as it is done one line below for the b-tagging.
Figure 3: do you have any clue why the Data/MC ratio is so close to 1 for photon events with respect to Z→mumu or Z→ee?
L198: here you define Delta R, why it should have been defined earlier (see comment for L145)
L393-395: I do not understand this statement
Figure 7: what is “EWK” and why is it so asymmetric in u_parallel for photon+jets events?
Table 2: why gamma+jets sigma_0 measurement has a much smaller error with respect to Z+jets?
Table 3: same comment as for Table 2 + in general there is some tension among Z+jets and photon+jets channels… maybe errors are underestimated?
Figure 13: Why Calo Et has a better data-MC agreement with respect to PF Et?
L516-519: it would be more readable after rephrasing like: “The following inputs are used for the BDT: …”
L527-528: “They agree…” I would drop this sentence, since it is not possible to get it from the plots. BTW: is it expected and desirable?
L564-565: why “variables”? Isn't it just Et the variable of interest?
L674-680: here the authors list as “interesting features” a couple of things which follow directly from the definition and the (correct) implementation of the significance. Maybe it would be better to rephrase with “It can be noticed how events arising from zero true Et … are mostly found at low values of significance…”
L714-716: What is instead the performance of the simple Et cut as a function of pile-up? Does it degrade more or less than the Et-significance?
Figure 29: I would put the reference to fig 22-23 in the caption rather than in the plot itself (which is also painful if you change somehow the order or the number of the figures in the document)
// ================================================================
Mario:
General style comments:
- Unsure if the title is proper English. Shouldn't it be “Performance … with the 8 TeV pp data” ? - We have in general some issues with the terminology used in the paper “Missing energy reconstruction”. We do not reconstruct the missing energy, we reconstruct everything else, and then we estimate/determine/evaluate the missing energy. - All figures report “CMS Preliminary”. If this is a paper, shouldn't the preliminary be dropped?
Line by line comments: line 2: “… detector can detect …” → “… experiment can detect…” Section 2: it is somewhat unsettling that the tracker is described last. Perhaps it would make more easy to follow to put it as a first detector described, using the usual convention of describing from inside-out. line 50: “… in two endcap regons…” → “in the two endcap regions…” line 80: “…associated to that vertex…” → “…associated…” line 94 “…associated with an energy deficit in the calorimeters.” sound very unclear to an uneducated reader not familiar with the CMS muon reconstruction. We propose to remove this part of the sentence. lines 177-182: it will be probably unclear to an external reader which part of the selection you are mentioning refers to the trigger selection and which part to the offline selection. Can you make clearer which are the trigger and which are the offline selection criteria? line 207: “…misidentification of jets with large electromagnetic fraction as photons.” → “…misidentification as photons of jets with large electromagnetic fraction.” line 219: utilizing → using line 242: it is somewhat incorrect to classify neutrinos from leptonic decays as overestimation of missing transverse energy. One should add a sentence explaining that we mostly care for neutrinos from the primary interaction, and identify as background neutrinos produced by secondary vertices line 252: “due to the calorimeter non-linearity and minimum energy thresholds in the calorimeters” is the first just ECAL (as opposed to ECAL+HCAL in the second usage of the word calorimeters)? In the second case, we suggest: “due to the non-linearity and minimum energy thresholds in the calorimeters” line 310: datastreams seems jargon. Perhaps “data samples”? lines 340-344: since E_T is a vectorial quantity, it is perhaps better to point out here that all these “sources of E_T” do not necessarily go in the same direction, and can even cancel out“ line 382: “and also symmetric” → “and to be also symmetric” Figure 9: there seem to be some different behavior in the Z and photon samples for the perpendicular recoil component (right figure) around 100 GeV. Is this normal? Can a sentence be added in the text to explain why this is happening? line 433: “in Z → µ+µ−, Z → e+e− , and photon events.” → “in events where a boson is present” line 472: ”…charged particles that are neither associated to leptons nor to jets of pT > 30 GeV…“ → ”…charged particles of pT > 30 GeV that are neither 473 associated to leptons nor to jets…“ line 531: “jet energy resolutions” → better use the singular “jet energy resolution … is… ” line 627: conducted → performed line 661: apparent → evident line 736: “non-Gaussian resolutions” → “non-Gaussian contributions to resolution”
// ================================================================
Gian Luca:
Commenti di carattere generale: Ho trovato spesso poco chiaro quando vengono usati dati “reali” o campioni MC.
da riga 9 a riga 15 non é un cosa “importante” ma che trovo comunque non corretta. Qui parla dell'importanza dell'uso della missing energy per canali di supersimmetria e in generale analisi oltre il modello standard e parla invece delle analisi come higgs, ZZ e WW ecc come importanti al passato e cioé come se fossero delle analisi finite. Invece proprio in questi canali si potrebbero trovare deviazioni dal modello standard come ad esempio la sezione d'urto differenziale in processi di VV scattering.
riga 20: C'é la fedinizione di pile up che non é un errore di per se peró non capisco perché spieghi il pile up e molte altre cose molto meno note e banali le dia per scontate.
riga 57: Cé una formula che all'inizio non mi é stata di chiara lettura ma perché non ho seguito il corso di calorimetria. Non conoscevo l'uso di ⊕ come somma in quadratura. Forse puó starci scriverlo.
riga 80: The symbol P_t is already defined, I suggest to simplify the sentence using it. (visto che ha giá definito il p_t come momento trasverso nella stessa pagina qui si potrebbe usare la per alleggerire la frase)
riga 88: Trovo la frase pesante. not linked to the extrapolation of any charge… → with no links between a charged-particle track.
riga 94: With an energy deficit..→ without a corresponding ammount of energy in the calorimeter.(o qualcosa del genere. il calorimetro non ha dei deficit. al massimo non ha dei depositi di energia)
riga 118: collision data sound strange. collision data→ real data or simply data.
riga 145: Delta R is used but is defined after in the section 3.4 line 198.
line 184: photons ad jets → photons and at least one jet
line 198: look line 145.
line 200: If in this selection particle flow is uesd the photon reconstruction is already described in line 88.
line 226: the sentence sounds confusing, I suggest to rephrase it.
section 7.3 : Is unclear when MC or real data are used. Is very very misleading.
// ================================================================
Linda:
10: standard model –> Standard Model
111-113: non trovo chiara la frase “The jet momentum … is found in simulated samples to be within 5% and 10% of the true momentum over the whole pT spectrum and detector acceptance”. Significa che il jet ha circa il 5-10% del pT totale?
205: se il soggetto e' “majority” il verbo non dovrebbe essere “fails” al posto di “fail”? (Ma forse mi sbaglio io!)
243-245: non mi pare che spieghi perche' “this bias is significantly reduced by correcting the pT of the jets to the particle-level pT using the jet energy corrections” (magari e' una cosa ovvia, ma io non l'ho capita)
248: In addition to the corrections –> In addition to the correction (mi pare sia solo quella del pT dei jet)
334: c'e' differenza tra “well-measured” e “well-understood”? Non e' ridondante?
391: non definisce “uncorrected unclustered energy”
462, 468: “MVA pileup jet ID” nella riga 450 era definita solo come “pileup jet ID”
473: “unclustered charged hadrons” in riga 463 erano “unclustered HS-charged hadrons”
499: forse e' banale, ma secondo me sarebbe meglio specificare se i jet che passano la “MVA pileup jet ID” siano jet di pileup oppure no.
541: “The No-PF ET response approches unity slower than standard PF ET” –> aggiungerei tra parentesi la figura di riferimento (fig 8)
562: its use –> the use
573: “A first version of THIS algorithm”, non si capisce quale perche' sopra non ne spiega uno
618, 625, 650…: Z→mu mu –> Z→mu+mu- (finora e' sempre stato scritto cosi')
714, 716 pile up –> pileup (finora e' sempre stato scritto cosi')
// ================================================================
Marta:
o Abstract The present version is too verbose and difficult to read. Please try to rephrase it more compact.
o Introduction - line 12: delete “at a mass of around 125 GeV” - line 21: “algorithms and corrections related to missing_transverse_momentum reconstruction, together with performance studies carried on 8 TeV …” - line 22: move sentence on the average number of pileup interactions at the beginning of Section 3, after the integrated lumi. - line 26: ”..simulation sample used for the present study…“ - line 27: “In Section 4, the different missing_transverse_momentum reconstruction algorithms are presented. In Section 5, the issue of missing_transverse_momentum misreconstruction is discussed, its sources are identified and corrections are addressed. ”
- line 31: “The degradation of the resolution due to pileup
interactions is studied in detail.”
o Section 3 - line 75: “The data sample for the present study…” - line 80: ”…of all tracks associated to it is considered to be the primary vertex, corresponding to the origin of the hard-scattering process.“ - line 88: “not linked” → “not matched” - line 89 ”…identified as primary charged particles trackS reconstructed by a Gaussian-sum filter (GSF) algorithm and are extrapolated and matched to ECAL energy clusters, possibly also including associated…“ - line 93: ”…identified as tracks in the central tracker matched with…“ - line 97: ”….energy clusters not matched to any charged…“ - line 100: “The energy of photons is directly obtained from the ECAL measurement”. (not necessary to mention systematic effects) - line 101: “The energy of electrons is determined from the corresponding ECAL cluster energy…” - line 111: delete “in this jet” - line 118: delete “that are” - line 119: “PYTHIA 6 tune Z2 [11,13] - line 120: delete sentence “PYTHIA 6…Z2 [13]” - line 123: ”…subsystem conditions among which bad channels.” - replace lines 125 to 132 with: “To account for pileup interactions the simulated vertex multiplicity (N_vtx) distribution is weighted to describe the data (see Fig. 1 for for the process Z → e+e-). The uncertainty on the N_vtx distribution hence affects the pileup profile in the simulation. It is dominated by the uncertainties on the total inelastic pp cross section measurement [15,16] and on the luminosity measurement. The latter contributes 30% of N_vtx uncertainty.” - after line 132 it would be good to add a short paragraph to introduce what follows (description of the different data samples used for the present study). One could say in this paragraph (rather than later on in sections 3.1, 3.2, 3.3, 3.4) which sample is used for what (e.g. the dijet sample used for anomalous high missing_transverse_momentum analysis).
o Section 3.1 - line 136, 137: delete the sentence “The integrated luminosity….bf-1.” - replace lines 137 to 139 with: “Dijet events are selected by requiring a leading jet with p_T > 400 GeV, which ensures full trigger efficiency, and at least another jet with p_T > 200 GeV.”
o Section 3.2 - line 144: “through hadron decay” - line 148: “…vertex, OF neutral hadrons and OF photons.” - lines 152-153: delete sentence “We require….criteria:” and start with “Muon candidates are required to be reconstructed in the tracker and in the muon chambers, TO have a reconstructed …20 GeV, and TO be…” - line 157: “The Z→e+e- events were…” - lines 158 to 164: delete sentence “The events are required…..criteria:” and go on with “Electron candidates are required to be within the ECAL …1.56 < |eta| < 2.5 and to have a reconstructed transverse momentum p_T greater than 20 GeV. To reject events…by satisfying the condition in Eq.(1)” - Eq. (1) is not a condition but just a definition of R_iso - lines 166-168: In Fig. 3 there are simulated samples not described in the text. Same comments applies to other figures later on.
o Section 3.3 - line 170: “The W→ e nu events were collected…” - define medium and loose working points before mentioning them. - lines 177-178: delete “from the full 2012 CMS dataset.”
o Section 3.4 - line 187: “The trigger p_T boundaries…” - line 191: “…to pass THE selection criteria…” - line 199: “…photon candidates produced inside…” - line 202: delete last sentence “Events respecting….sample.” - line 207: “…and TO THE occasional….” - line 211-212: merge the two paragraphs (topic is the same) - line 212: delete first sentence “We utilize…variable.” and go on with “A sample of collision data events is identified where…requirements.” - line 215: delete “From the distribution in the sample we remove” and go on with “Non-QCD background processes are removed by subtracting their simulated distributions. - lines 216-217 delete “as estimated by simulation. The remaining …background.” - lines 217-219 are not clear
o Section 4 - First sentence to be rephrased, e.g., though not perfect yet “Missing_transverse_momentum is the imbalance in the transverse momentum of all MEASURED particles in the final states, hence interacting via the electromagnetic or strong forces” - line 239: ”…mitigate THE effects…“ - line 243: ”…the response of HCAL due to its non-compensating nature.“ - line 244: ”…by correcting the jet p_T….“ - line 248: “Further corrections improve the performance…” - line 252: “However, due to non linearities and minimum energy thresholds of the calorimeters, …” - line 256-257: merge the two paragraphs (topic is the same) - line 267: “The sources of such asymmetry are related to imperfect…” - line 270: ”…not only in THE collision data but also…“ - line 272: ”…denoted BY…“ - line 293: ”…are discussed in Ref.[22]“ - line 297: ”…low-quality hits and are removed.“ - line 301: ”…the channels used for signal…“ - line 306:”…some EVENTS are read out and reconstructed. The transverse momentum of reconstructed fake tracks can exceed 100 GeV, thus miming charged particles, which are then clustered in jets with high p_T and produce strip and pixel …“ - line 314: “A noise-rejection algorithm was developed to exploit the differences between noise and signal pulse shapes.” - line 317: ”…pulse shape: measured and expected signal pulse shapes are compared and several…“ - line 319: “Figure 4 shows the resulting…for the dijet sample described in Sec. 3.1. - line 322: “Even after the application of all the …2012 data, we still find a small ….Imposing jet identification criteria to limit the maximum …., guarantees …” - line 329: ”…and agrees well with the simulated distributions (Fig. 4).”
o Section 6 - line 333: refer to paragraph 3.2, 3.4 where the data samples are described. - line 335: “system, which dominates…” - lines 340 to 344: “Even if no genuine…is expected in a process, many physics (pileup, underlying event activity, fluctuation in jet composition) and detector (noise, particle misreconstruction, energy resolution, jet energy corrections) effects can significantly affect the missing_transverse_momentum measurement. - line 348: ”…while the jet energy..“ - line 349-350: ” …dominated by the contribution of the hadronic activity in the event.“ - line 358: ”…is related to the jet energy resolution uncertainty in…“ - lines 369-371: ”…in the hemisphere opposite to the boson…“
o Section 6.1 - line 379: ”…provides a natural basis on which to evaluate…“ - line 393: ”..lower than the fit to Z data…“ - line 399-400: ”..spectra are different, which implies differences in the corresponding resolution curves.“ - line 402: ”…100 GeV. An event-by-event…“ - line 411: ”..and is consistent for different channels as shown…“ - line 414: ”…dependence on N_vtx…“ - line 426: ”…are shown on Fig. 12 and 13.“
o Section 7.1 - lines 452-455: to be moved after line 442 in previous session
o Section 7.3 - line 499: ”…and OF all neutral PF particles…“ - line 502: ”…and OF all neutral PF particles…“ - line 505: ”…and OF all neutral PF particles…“ - line 571: ”…THE scalar sum of transverse momenta of all PF particles of each corresponding…“
o Section 8.2 ”…general broadening OF the…“
o Section 8.4 - line 698: ”..but we find that THE performance…“ - line 699: ”
o Section 9 - line 753: “The data sample used for this study…”
o Figure 2: “produced in THE final state” (typing error)
o Figure 3: “The error bars on the data/MC ratio plots represent the statistical uncertainty, whereas the grey error band shows the systematic uncertainty on the simulation.”
o Figure 5: “The error bars on the data/MC ratio plots represent the statistical uncertainty, whereas the grey error band shows the systematic uncertainty on the simulation.”
o Figure 7: “Grey bands represent the statistical and systematic uncertainties on the simulation.”
o Figure 8 and 9, last sentence: “…value in THE data contributing to each point.”
o Figure 13: “…Z→nu=nu- data events.”
o Figure 18, 19, 20: “compoNent” (typing error)
o Table 3: “…propagation into the parameterization of the following effects:…”