From womersley@fnal.gov Fri Mar 30 11:35:51 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2UHZpD30336 for ; Fri, 30 Mar 2001 11:35:51 -0600 Received: from fnal.gov ([131.225.231.26]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB0001KVU7PEX@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Fri, 30 Mar 2001 11:35:50 -0600 (CST) Date: Fri, 30 Mar 2001 11:27:56 -0600 From: John Womersley Subject: Re: Photon Ratio PRL Draft Available To: Mike Strauss Cc: chopra@fnal.gov, dhiman@fnal.gov, jkrane@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Message-id: <3AC4C21C.30867BDD@fnal.gov> Organization: Fermi National Accelerator Laboratory MIME-version: 1.0 X-Mailer: Mozilla 4.74 [en] (WinNT; U) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200103292250.f2TModq05750@particle.nhn.ou.edu> Content-Length: 1797 Status: RO Mike, some quick comments, nothing serious: a) in Ref [6] you can also cite Walter Giele, Proc. of the 5th Int. Symposium on Radiative Corrections, Carmel, CA, September 2000. To be published electronically at http://www.slac.stanford.edu/econf/ b) Fig.4. should make clear that the cross section is at 630 GeV. I think we may want to change the way we present the systematic errors in this plot, even though what you have is some kind of D0 convention. As an example, at the lowest CC point, the figure implies that the data point is many sigma of systematic error above theory. In fact, it is less than 2 sigma off. This can be figured out in the table, but... c) Fig.5. presentation of theory should be same as Fig.4. i.e. a solid line for Owens. Would be nice to get Vogelsang too. d) when you calculate the chi squared, do you use the "D0 recipe" of renormalizing the systematic errors by the theory rather than the data? (I think to do so could be wrong in this case given that the sys error is dominated by purity). Have you contacted the theorists? John -------------------------------------------------------------------- From strauss@phyast.nhn.ou.edu Fri Mar 30 13:57:13 2001 Return-Path: Received: from particle.nhn.ou.edu (particle.nhn.ou.edu [129.15.30.205]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2UJtBD29516; Fri, 30 Mar 2001 13:55:11 -0600 Received: by particle.nhn.ou.edu (8.10.2/8.10.2) id f2UJt9606994; Fri, 30 Mar 2001 13:55:09 -0600 (CST) Date: Fri, 30 Mar 2001 13:55:09 -0600 (CST) From: Mike Strauss Message-Id: <200103301955.f2UJt9606994@particle.nhn.ou.edu> To: strauss@mail.nhn.ou.edu, womersley@fnal.gov Subject: Re: Photon Ratio PRL Draft Available Cc: chopra@fnal.gov, dhiman@fnal.gov, jkrane@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: 7i8yQs8fZK4lFjSGyWaXFA== Content-Length: 2877 Status: RO John, Thanks for your comments. Here are some responses. > Mike, > > some quick comments, nothing serious: > > a) in Ref [6] you can also cite Walter Giele, Proc. of the > 5th Int. Symposium on Radiative Corrections, Carmel, CA, > September 2000. To be published electronically at > http://www.slac.stanford.edu/econf/ Fine. I will do this. > b) Fig.4. should make clear that the cross section is at 630 GeV. I will add something in the caption of figures 3 and 4 to state that the results are at 630 GeV. (Would it be better in the figure itself?) > I think we may want to change the way we present the systematic > errors in this plot, even though what you have is some kind of > D0 convention. As an example, at the lowest CC point, the > figure implies that the data point is many sigma of systematic > error above theory. In fact, it is less than 2 sigma off. > This can be figured out in the table, but... I would like to discuss this point more. When I calculate the chi^2 value, I multiply the percent correlated error by the theory, not the data. So the way it is presented in the plot is actually consistent with what is done. In addition, the largest correlated error for the CC (and the low E_T of the EC) is in the fragmentation model, which is not associated with the data at all, and is in some sense more of a theoretical error. So, for now, the figure is consistent with what is done and maybe even correct. Any further thoughts or thoughts from others? > c) Fig.5. presentation of theory should be same as Fig.4. i.e.> a solid line for Owens. Would be nice to get Vogelsang too. I have contacted Owens and Vogelsang and am working on this. When I made the figure initally, I only had 1800 predictions at the central E_T value. Thus I created the plot shown. When I get 1800 in 1 GeV bins, I can change this. > d) when you calculate the chi squared, do you use the "D0 recipe" > of renormalizing the systematic errors by the theory rather > than the data? (I think to do so could be wrong in this case > given that the sys error is dominated by purity). The correlated errors are normalized by the theory, but the uncorrelated errors associated with statistics (e.g. the Purity) are done by renormalization of the data. There are really no systematic errors, but uncorrelated (predominately data and statistically driven, dominated by the purity, and normalized by the data), and correlated (most of which are not driven by statistics and data, dominated by fragmentation for CC, and low E_T for EC, and normalized by the theory). The EB has discussed this, but I am certainly open to changing this if there is a "correct" way of doing it. > Have you contacted the theorists? > Yes. Owens and Vogelsang are in touch with me, and with each other. -Mike -------------------------------------------------------------------- From jkrane@fnal.gov Fri Mar 30 14:38:27 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2UKcQD23080 for ; Fri, 30 Mar 2001 14:38:26 -0600 Received: from fnal.gov ([131.225.226.198]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB100CKS2M492@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Fri, 30 Mar 2001 14:37:16 -0600 (CST) Date: Fri, 30 Mar 2001 14:38:14 -0600 From: John Krane Subject: Re: Photon Ratio PRL Draft Available To: Mike Strauss Cc: womersley@fnal.gov, chopra@fnal.gov, dhiman@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Message-id: <3AC4EEB6.D4620456@fnal.gov> Organization: Iowa State University MIME-version: 1.0 X-Mailer: Mozilla 4.76 [en] (WinNT; U) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200103301955.f2UJt9606994@particle.nhn.ou.edu> Content-Length: 667 Status: RO Hi everyone, The treatment of the systematics is defensible and (I assert) correct. I'm willing to discuss this more in person, but don't have time over email. Bertram could discuss also of course. Anyway, Figure 4 is correct in this view, even though the error is mostly purity instead of energy scale or some other systematic. The corrolary is, Fig. 5 should probably be updated to put the errors on the theory line-which-is-currently-points. Although the current Fig. 5 is consistent with my PRL, that style was updated in the PRD to be consistent with what we actually do (i.e. your Fig. 4). - John -- http://www-d0.fnal.gov/~jkrane/John_Krane.html --------------------------------------------------------------------- From womersley@fnal.gov Fri Mar 30 17:29:40 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2UNTeD13244 for ; Fri, 30 Mar 2001 17:29:40 -0600 Received: from fnal.gov ([131.225.231.74]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB100JOAALFWU@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Fri, 30 Mar 2001 17:29:39 -0600 (CST) Date: Fri, 30 Mar 2001 17:29:39 -0600 From: "Captain's account" Subject: Re: Photon Ratio PRL Draft Available Sender: d0cap@fnal.gov To: John Krane Cc: Mike Strauss , chopra@fnal.gov, dhiman@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Message-id: <3AC516E3.4EB15F3C@fnal.gov> MIME-version: 1.0 X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.2.16-3smp i686) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200103301955.f2UJt9606994@particle.nhn.ou.edu> <3AC4EEB6.D4620456@fnal.gov> Content-Length: 1300 Status: RO Mike, John et al., I think that what you have done in defining chisquared conforms to D0 practice and is certainly defensible. (I mis-spoke when I brought purity into the question: the purity errors are uncorrelated). It should be made clear in the paper that chi squared is defined this way. I understand why we want to use this definition, but I just want to make sure we are all aware of what this means and comfortable with its application when the data is four times the theory --- it is not a minor adjustment to the chi squared in this case! John John Krane wrote: > Hi everyone, > > The treatment of the systematics is defensible and (I assert) > correct. I'm willing to discuss this more in person, but don't > have time over email. Bertram could discuss also of course. > Anyway, Figure 4 is correct in this view, even though the error > is mostly purity instead of energy scale or some other > systematic. > > The corrolary is, Fig. 5 should probably be updated to put the > errors on the theory line-which-is-currently-points. Although > the current Fig. 5 is consistent with my PRL, that style was > updated in the PRD to be consistent with what we actually do > (i.e. your Fig. 4). > > - John > > -- > > http://www-d0.fnal.gov/~jkrane/John_Krane.html ------------------------------------------------------------------- From strauss@phyast.nhn.ou.edu Fri Mar 30 15:34:07 2001 Return-Path: Received: from particle.nhn.ou.edu (particle.nhn.ou.edu [129.15.30.205]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2ULWnD28446; Fri, 30 Mar 2001 15:32:49 -0600 Received: by particle.nhn.ou.edu (8.10.2/8.10.2) id f2ULWlQ07083; Fri, 30 Mar 2001 15:32:48 -0600 (CST) Date: Fri, 30 Mar 2001 15:32:48 -0600 (CST) From: Mike Strauss Message-Id: <200103302132.f2ULWlQ07083@particle.nhn.ou.edu> To: strauss@mail.nhn.ou.edu, jkrane@fnal.gov Subject: Re: Photon Ratio PRL Draft Available Cc: womersley@fnal.gov, chopra@fnal.gov, dhiman@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: UQYVi7y12KUW1XIbQG8ztQ== Content-Length: 1130 Status: RO I have a question/response to the second part of John's comments. > Hi everyone, > > The corrolary is, Fig. 5 should probably be updated to put the > errors on the theory line-which-is-currently-points. Although > the current Fig. 5 is consistent with my PRL, that style was > updated in the PRD to be consistent with what we actually do > (i.e. your Fig. 4). There is a problem with making figure 5 look like Figure 4. The correlated errors in Figure 5 are calculated in each bin because I am using the correlated errors from the 1800 PRL. That means I do not really have a smooth curve (or 1 GeV bins) for the correlated error. Therefore, the most straightforward way to display the correlated error is bin by bin, done exactly like the Jet PRL and done the way that the correlated error is actually calculated in the ratio. It may be possible to get the correlated errors done in 1 GeV bins for the 1800 data but I'm not sure we gain anything except asthetics? Does Steve know how easy (hard? possible?) it is to get the correlated errors in 1 GeV bins from the 1800 measurement? -Mike > - John > > ----------------------------------------------------------------------- From strauss@phyast.nhn.ou.edu Fri Mar 30 15:42:16 2001 Return-Path: Received: from particle.nhn.ou.edu (particle.nhn.ou.edu [129.15.30.205]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2ULerD31098; Fri, 30 Mar 2001 15:40:53 -0600 Received: by particle.nhn.ou.edu (8.10.2/8.10.2) id f2ULeqZ07091; Fri, 30 Mar 2001 15:40:52 -0600 (CST) Date: Fri, 30 Mar 2001 15:40:52 -0600 (CST) From: Mike Strauss Message-Id: <200103302140.f2ULeqZ07091@particle.nhn.ou.edu> To: strauss@mail.nhn.ou.edu, jkrane@fnal.gov Subject: Re: Photon Ratio PRL Draft Available Cc: womersley@fnal.gov, chopra@fnal.gov, dhiman@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: /kd2hvpFoYpvDMCNwgvZgQ== Content-Length: 758 Status: RO One more comment about this. I could put correlated "boxes" for each bin around the theory rather than around the data. As John states, that is consistent with what we actually do. But a correlated "line" is not really consistent with what we do, and is much harder to extract from the 1800 GeV analysis. If we change the points to a line as Womersley sugests, then it might be necessary to get the correlated uncertainty in 1 GeV bins at 1800. > The corrolary is, Fig. 5 should probably be updated to put the > errors on the theory line-which-is-currently-points. Although > the current Fig. 5 is consistent with my PRL, that style was > updated in the PRD to be consistent with what we actually do > (i.e. your Fig. 4). > > - John > -Mike --------------------------------------------------------------------- From jkrane@fnal.gov Fri Mar 30 15:43:40 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2ULheD15260 for ; Fri, 30 Mar 2001 15:43:40 -0600 Received: from fnal.gov ([131.225.226.198]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB100L1Q5ORJU@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Fri, 30 Mar 2001 15:43:39 -0600 (CST) Date: Fri, 30 Mar 2001 15:44:38 -0600 From: John Krane Subject: Re: Photon Ratio PRL Draft Available To: Mike Strauss Cc: womersley@fnal.gov, chopra@fnal.gov, dhiman@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, daniel@fnal.gov, linn@scri.fsu.edu, marek@d0mino.fnal.gov Message-id: <3AC4FE46.32D4DFC1@fnal.gov> Organization: Iowa State University MIME-version: 1.0 X-Mailer: Mozilla 4.76 [en] (WinNT; U) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200103302132.f2ULWlQ07083@particle.nhn.ou.edu> Content-Length: 198 Status: RO Mike, what I meant was the errors should be err->err/data*theory, and plotted on the theory points. I'm not talking about rebinning. - John -- http://www-d0.fnal.gov/~jkrane/John_Krane.html ------------------------------------------------------------------------ From jkrane@fnal.gov Fri Mar 30 15:44:27 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f2ULiRD24818 for ; Fri, 30 Mar 2001 15:44:27 -0600 Received: from fnal.gov ([131.225.226.198]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB100K6B5Q2DW@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Fri, 30 Mar 2001 15:44:26 -0600 (CST) Date: Fri, 30 Mar 2001 15:45:25 -0600 From: John Krane Subject: Re: Photon Ratio PRL Draft Available To: Mike Strauss Message-id: <3AC4FE75.931E46CF@fnal.gov> Organization: Iowa State University MIME-version: 1.0 X-Mailer: Mozilla 4.76 [en] (WinNT; U) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200103302140.f2ULeqZ07091@particle.nhn.ou.edu> Content-Length: 181 Status: RO Oh, and just fit the theory points with a polynominal or something to get a line. It's only for display purposes. - John -- http://www-d0.fnal.gov/~jkrane/John_Krane.html ------------------------------------------------------------------------- From jkrane@fnal.gov Tue Apr 3 15:17:52 2001 Return-Path: Received: from fnal.gov (heffalump.fnal.gov [131.225.9.20]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f33KHqD29874 for ; Tue, 3 Apr 2001 15:17:52 -0500 Received: from fnal.gov ([131.225.226.198]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GB800KHGGDRRZ@smtp.fnal.gov> for strauss@mail.nhn.ou.edu; Tue, 03 Apr 2001 15:17:51 -0500 (CDT) Date: Tue, 03 Apr 2001 15:18:58 -0500 From: John Krane Subject: comments on initial version To: Mike Strauss Cc: Marek Zielinski Message-id: <3ACA3032.D2CF5F02@fnal.gov> Organization: Iowa State University MIME-version: 1.0 X-Mailer: Mozilla 4.76 [en] (WinNT; U) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: <200104031955.f33JtlN09036@particle.nhn.ou.edu> Content-Length: 1644 Status: RO Hi Mike, I'm faxing the marked-up version to you as I hit send on this email. You will find many many syntax suggestions from me, all trying to improve word flow where my ear detects a log jam. Mostly, I'm trying to eliminate passive voice (where the subject of the sentence is having something *done to it* instead of doing something). In general, a reader can more easily understand the active voice, so the article seems more clear as a result. You should feel free to incorporate as much or a little as you like. Please do not interpret my many marks as an insult; I know how difficult it can be to write these articles...and how difficult it can be to get a marked-up copy back... You should look out for my comments in the margins and white spaces. In spots, I think you need to say more, in others less. One thing I noticed was your changing of verb tense. You need to pick one (I vote for present tense) and stick to it. Things worth highlighting: - you should reference the PDFs - you need to make sure all fonts in the pictures are 2mm or larger in 2 column mode - the table columns should be renamed (they are currently unclear) - the sys error blocks should be put on the theory in Fig 5 - computing the data-theory chi2 is not the same as performing a fit Finally, after looking at the 2 column version, I see you have approximately one column too many for PRL, so there is some trimming you can do. Some of this can be accomplished by relying more heavily on the 1800 paper ("the procedure for blah-blah follows the methods outlined in [ref]..."). - John -- http://www-d0.fnal.gov/~jkrane/John_Krane.html ---------------------------------------------------------------------- From daniel@fnal.gov Thu Apr 19 17:46:31 2001 Return-Path: Received: from d0mino.fnal.gov (d0mino.fnal.gov [131.225.224.45]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f3JMkVl28512 for ; Thu, 19 Apr 2001 17:46:31 -0500 Received: from localhost (daniel@localhost) by d0mino.fnal.gov (SGI-8.9.3/8.9.3) with ESMTP id RAA89065 for ; Thu, 19 Apr 2001 17:46:30 -0500 (CDT) X-Authentication-Warning: d0mino.fnal.gov: daniel owned process doing -bs Date: Thu, 19 Apr 2001 17:46:30 -0500 From: "V. Daniel Elvira" To: Mike Strauss Subject: Re: Photon Ratio PRL Draft Available In-Reply-To: <200104192242.f3JMgN322670@particle.nhn.ou.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 1044 Status: RO Thanks. I understand the you get the theory curves directly from the theorists (we don't have the capability of running it ourselves). Do we have any feeling about the theory depencence on PDFs and muscales? (the uncertainty in the theory). Daniel ------------------------------------------------------------------------------- Daniel Elvira PAT/CD Fermilab daniel@fnal.gov (630)840-3604 ------------------------------------------------------------------------------- On Thu, 19 Apr 2001, Mike Strauss wrote: > John Krane has a number of comments which are being incorporated into > a second draft. Also, John thinks it is about one single-spaced > column too long. Finally, Werner Vogelsang is still working on some > of the theoretical calculations along with Jeff Owens. They are > trying to iron out a few differences in results that they get. > Once I get the final results from them and incorporate John's comments > we should send it to the Run I group. Other members of the EB have > not made comments. > > -Mike > ------------------------------------------------------------------------ From mont@d0mino.fnal.gov Thu Apr 19 23:36:17 2001 Return-Path: Received: from d0mino.fnal.gov (d0mino.fnal.gov [131.225.224.45]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f3K4aGl06700 for ; Thu, 19 Apr 2001 23:36:16 -0500 Received: from localhost (mont@localhost) by d0mino.fnal.gov (SGI-8.9.3/8.9.3) with SMTP id XAA20475; Thu, 19 Apr 2001 23:36:15 -0500 (CDT) Message-Id: <200104200436.XAA20475@d0mino.fnal.gov> X-Authentication-Warning: d0mino.fnal.gov: mont@localhost didn't use HELO protocol X-Mailer: exmh version 2.0.2 2/24/98 From: Hugh Montgomery Organization: Fermi National Accelerator Laboratories Precedence: normal To: "Mike Strauss" , chopra@fnal.gov, dhiman@fnal.gov, jkrane@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, qianj@umich.edu, womersley@fnal.gov, daniel@fnal.gov, linn@scri.fsu.edu Cc: mont@fnal.gov Subject: 630 GeV Photons Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 19 Apr 2001 23:36:15 CDT Sender: mont@d0mino.fnal.gov Content-Length: 2234 Status: RO Mike, Jianming pointed me to your paper draft. I couldn't resist. I am very pleased to see it appearing. Comments below. ( They are all relatively minor.) Mont ========================================================================= o page 8, para 1, the absence of braces makes the resolution formula potentially ambiguous. o page 8, para 3, second sentence "Geometric..." is superfluous given the third sentence "Clusters are required to lie.." o page 8, para 3, change ".. 50 cm from .." to ".. 50 cm of .." o page 9, para 1, the mention of the presence of "pions" in the test beam could confuse the reader... he might start to think pizero and gammas from pizero.. and while we tried that, I don't believe that is used .. o page 9, para 1, there is a sentence with "reject" and "rejected" both. rephrase to avoid repetition. o page 9, para 1, change "This selection criteria.." to "This selection criterion.." o page 9, para 2, the formula log... etc... Wow!, do we need to explain why we use this particular form? o page 11, para 3, when we mention Fig. 3, we should probably mention the curves therein. o page 11, para 4, I was looking for a mention of the figure in the first line, it does come but only in third sentence. Move up? o page 16, para 1, I am not sure I understand why Vogelsang definese E_T^max using the "largest" photon energy. How many does he expect?? o page 16, para 1, after you are done, and despite the "acceptable" chisqd comparison, I am tempted to mention that the agreement is better at high p_T or worse at low p_T. o page 16, para 3, change "The data was fit..." to "The data were fit..." o page 17, para 1... are you not tempted to say taht the ratio of cross sections is better described by the theory than the cross sections themselves? I understand that with the QCD group discovering chisqd, it has become quite circumspect in its conclusions (rightly)... nevertheless, what you see fits my prejudice.. so I would probably have suggested taht the data mildly support the idea that the theorists do better when pdfs play a smaller role. ------------------------------------------------------------------------- From strauss@phyast.nhn.ou.edu Tue May 22 16:33:17 2001 Return-Path: Received: from particle.nhn.ou.edu (particle.nhn.ou.edu [129.15.30.205]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f4MLXH718014; Tue, 22 May 2001 16:33:17 -0500 Received: by particle.nhn.ou.edu (8.10.2/8.10.2) id f4MLXFi06216; Tue, 22 May 2001 16:33:15 -0500 (CDT) Date: Tue, 22 May 2001 16:33:15 -0500 (CDT) From: Mike Strauss Message-Id: <200105222133.f4MLXFi06216@particle.nhn.ou.edu> To: marek@fnal.gov, mont@fnal.gov Subject: Responses to Mont on PRL Cc: strauss@mail.nhn.ou.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: v7YhuQQIOjYNyv4Db8+57Q== Content-Length: 3321 Status: RO > Mike, > > Jianming pointed me to your paper draft. I couldn't resist. I am very > pleased to see it appearing. > > Comments below. ( They are all relatively minor.) > > Mont > ========================================================================= > > o page 8, para 1, the absence of braces makes the resolution formula > potentially ambiguous. Braces added. > o page 8, para 3, second sentence "Geometric..." is superfluous given the > third sentence "Clusters are required to lie.." Agree. I deleted that whole sentence. > o page 8, para 3, change ".. 50 cm from .." to ".. 50 cm of .." Agree. Done. > o page 9, para 1, the mention of the presence of "pions" in the test beam > could confuse the reader... he might start to think pizero and gammas > from pizero.. and while we tried that, I don't believe that is used .. Agreed. I changed it to "charged pions". > o page 9, para 1, there is a sentence with "reject" and "rejected" both. > rephrase to avoid repetition. Changed second reference to "can discard" > o page 9, para 1, change "This selection criteria.." > to "This selection criterion.." Done > o page 9, para 2, the formula log... etc... Wow!, do we > need to explain why we use this particular form? I haven't explained it in detail. Just stated that it "emphasized the difference" between single photon showers and multiple photon showers. It is the same forumla used in the published 1800 paper. > o page 11, para 3, when we mention Fig. 3, we should probably mention the > curves therein. I still haven't done this. It is mentioned in the caption. > o page 11, para 4, I was looking for a mention of the figure in the first > line, it does come but only in third sentence. Move up? I haven't moved this up. It seems more natural to me in the third sentence since we need to mention the parameters used in the theory. > o page 16, para 1, I am not sure I understand why Vogelsang definese > E_T^max using the "largest" photon energy. How many does he expect?? The definition that Vogelsang uses was copied from the 1800 paper, but it was incorrect. I have contacted Vogelsang and now have the correct definition used. > o page 16, para 1, after you are done, and despite the "acceptable" chisqd > comparison, I am tempted to mention that the agreement is better at high > p_T or worse at low p_T. I have now done this. (Tom mentioned this, as well). > o page 16, para 3, change "The data was fit..." > to "The data were fit..." Actually, it is not fit, but compared. I have changed the wording. > > o page 17, para 1... are you not tempted to say taht the ratio of cross > sections is better described by the theory than the cross sections > themselves? I understand that with the QCD group discovering chisqd, > it has become quite circumspect in its conclusions (rightly)... > nevertheless, what you see fits my prejudice.. so I would probably have > suggested taht the data mildly support the idea that the theorists do > better when pdfs play a smaller role. > > I think the agreement in both is quite comparable. With large errors, I am reluctant to make too many strong conclusions or inferences. > > > -Mike ----------------------------------------------------------------------- From marek@d0mino.fnal.gov Tue May 22 16:50:48 2001 Return-Path: Received: from d0mino.fnal.gov (d0mino.fnal.gov [131.225.224.45]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f4MLom725076 for ; Tue, 22 May 2001 16:50:48 -0500 Received: from localhost (marek@localhost) by d0mino.fnal.gov (SGI-8.9.3/8.9.3) with SMTP id QAA43136 for ; Tue, 22 May 2001 16:50:46 -0500 (CDT) Message-Id: <200105222150.QAA43136@d0mino.fnal.gov> X-Authentication-Warning: d0mino.fnal.gov: marek@localhost didn't use HELO protocol X-Mailer: exmh version 2.0.2 2/24/98 To: strauss@mail.nhn.ou.edu Subject: here it comes again... Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 22 May 2001 16:50:46 CDT From: Marek Zielinski Content-Length: 5715 Status: RO ------- Forwarded Message Received: from heffalump (heffalump.fnal.gov [131.225.9.20]) by d0mino.fnal.gov (SGI-8.9.3/8.9.3) with SMTP id RAA25482 for ; Thu, 26 Apr 2001 17:25:03 -0500 (CDT) Received: from d0mino.fnal.gov ([131.225.224.45]) by smtp.fnal.gov (PMDF V6.0-24 #37519) with ESMTP id <0GCF00MWX7LQLQ@smtp.fnal.gov> for marek@d0mino.fnal.gov (ORCPT marek@fnal.gov); Thu, 26 Apr 2001 17:25:02 -0500 (CDT) Received: from localhost (marek@localhost) by d0mino.fnal.gov (SGI-8.9.3/8.9.3) with SMTP id RAA28575; Thu, 26 Apr 2001 17:25:02 -0500 (CDT) Date: Thu, 26 Apr 2001 17:25:02 -0500 (CDT) From: Marek Zielinski Subject: comments on the 630 Gev direct photon PRL draft To: chopra@fnal.gov, dhiman@fnal.gov, jkrane@fnal.gov, marek@fnal.gov, yasuda@fnal.gov, strauss@phyast.nhn.ou.edu, Steve Linn , John Womersley , qianj@umich.edu, "V. Daniel Elvira" Message-id: <200104262225.RAA28575@d0mino.fnal.gov> MIME-version: 1.0 X-Mailer: exmh version 2.0.2 2/24/98 Content-type: text/plain; charset=us-ascii X-Authentication-warning: d0mino.fnal.gov: marek@localhost didn't use HELO protocol Hi Mike. I am very pleased seeing your PRL taking shape! Nonetheless, here are some more (pesky) comments on the draft... Sorry for delayed response, but it's been rather hectic around here lately. I refer to the layout of the PRL-style version, dated 3/30/01. I will skip overlaps with previous comments by Mont. - - abstract, last sentence: Results for cross section at \sqrt(s)= 630 GeV and for .... in satisfactory agreement for most ot the p_T range. - - p.4, par 1: "... at the Tevatron..." is somewhat general; true at pt values you measured, but probably not for Run 2. Last sentence: remove "and fragmentation"? Jet fragmentation is the dominant background, and for signal, even after isolation cuts, fragmentation contribution is a difficult part for the theory. - - par 2, last word: function -> functions. - - par.3: expression for dimensionless cross section -- you could give it a name (say $\Sigma$) and refer later in the text and figure captions. Note that later you give different-looking formula for the same thing, more than once. Last sentence: seems stretched. True for D0/CDF, but not for direct photon measurements in general (unless you have a really good lawyer). - - end of par. 5: just checking that the 7.35 is indeed the number... - - caption to Fig. 1: " neutral meson background with charged tracks .... without charged" may be cryptic to someone not familiar with the analysis; neutral mesons do not leave tracks after all... The 1800 GeV paper used "jets with/without..." - - p5, par 1: "enhances" -> exploits, emphasizes? - - p5, par 2: "... overlaid with minimum bias..."? The 1800 GeV PRL states "overlaid with data acquired using a random trigger to model noise, pileup, underlying event, and multiple pbarp interactions" -- this implies zero-bias rather than min bias trigger, and gives a more expanded list of effects being taken care of. BTW, was underlying event switched off in Pythia, or are we slightly double counting? (or the UE part is not stated correctly?) Next sentence: "... neutral background events..." is our jargon, again. The events were not likely truly neutral; you could move here the phrase used 2 paragraphs later: "background with and without charged tracks in the road pointing back to the interaction vertex." Last sentence: "... match the data....from the photon data." does not read well. Also, without explanation, may invite questions (in readers mind) how we use low-purity photon DATA for this purpose. - - p5, last full par: you list components of uncertainties here, and again on p.6; is this needed? Also, you don't mention comparisons to theory in Fig. 3. - - p.6, par 1: "... where E_T^max is the maximum photon transverse energy in the event." The 1800 GeV paper states "E_T^max is the larger of the transverse energies of the photon or the leading jet." Which is right? " ... data was FIT to the theoretical cross section" -- I would prefer FIT-> compared "probability that the data fits the theory" -> "probability that the theory fits the data"? - - p.6, par 2: expression for dimesionless cross section is different than previously, could use a predefined symbol. "The ratio of the dimensionless cross sections (roots=630/roots=1800)" is likely understandable, but it is NOT the ratio of roots. If you decide to introduce a symbol for the lengthy formula, you could write more correctly $\Sigma(roots=630)/\Sigma(roots=1800)" (yes, picky,picky) - - p.6, par 3: duplicate list of contributions to uncertainties. Could be more useful to comment on the level of cancellations between various 630 and 1800 GeV systematics. - - p.7, caption to Fig:5: long formula, and "roots=630/roots=1800" comments as before. Add: vs x_T? Fig. 5, Vertical label "630/1800 Scaled" -> Dimensionless (or use symbols) - - p.7, par 1: "the data was fit to the theoretical" -> "the theoretical ... was fit to the data". Again "... probability that theory fits the data" "the overall results" -> "at higher x_T the results" In the 1800 GeV paper we decided not to brush aside the possibly interesting disagreement between NLO QCD and data at low pt! - - p.7, par 2, last sentence: "... and the theoretical prediction at higher E_T values" (same reason). Hope this helps. Keep us informed about new versions/plots/developments. Cheers, Marek ------- End of Forwarded Message ----------------------------------------------------------------------- From strauss@phyast.nhn.ou.edu Wed May 23 09:04:45 2001 Return-Path: Received: from particle.nhn.ou.edu (particle.nhn.ou.edu [129.15.30.205]) by phyast.nhn.ou.edu (8.10.2/8.10.2) with ESMTP id f4NE4j722106; Wed, 23 May 2001 09:04:45 -0500 Received: by particle.nhn.ou.edu (8.10.2/8.10.2) id f4NE4ir06649; Wed, 23 May 2001 09:04:44 -0500 (CDT) Date: Wed, 23 May 2001 09:04:44 -0500 (CDT) From: Mike Strauss Message-Id: <200105231404.f4NE4ir06649@particle.nhn.ou.edu> To: marek@d0mino.fnal.gov Subject: Comments from Markek, 26-apr Cc: strauss@mail.nhn.ou.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: OQ6vekHYxlMN+c0rhqR5Lw== Content-Length: 5844 Status: RO > Hi Mike. > > I am very pleased seeing your PRL taking shape! Nonetheless, here are > some more (pesky) comments on the draft... Sorry for delayed response, > but it's been rather hectic around here lately. I refer to the layout of > the PRL-style version, dated 3/30/01. I will skip overlaps with previous > comments by Mont. > > - - abstract, last sentence: > Results for cross section at \sqrt(s)= 630 GeV and for .... in > satisfactory agreement for most ot the p_T range. Done > - - p.4, par 1: "... at the Tevatron..." is somewhat general; true at pt > values you measured, but probably not for Run 2. > Last sentence: remove "and fragmentation"? Jet fragmentation > is the dominant background, and for signal, even after isolation cuts, > fragmentation contribution is a difficult part for the theory. Changed to "for modest E_T at the Tevatron" Ferbel's suggeston. I have not yet removed fragmentation. My thought is that the photon cross section still eliminates some of the fragmentation complications. > - - par 2, last word: function -> functions. Done > > - - par.3: expression for dimensionless cross section -- you could give it > a name (say $\Sigma$) and refer later in the text and figure captions. > Note that later you give different-looking formula for the same thing, > more than once. I give it a name later, when I define the experimental cross section. > Last sentence: seems stretched. True for D0/CDF, but not for direct > photon measurements in general (unless you have a really good lawyer). > Agreed. Sentence has been toned down. > - - end of par. 5: just checking that the 7.35 is indeed the number... It is. It was chosen to match 1800 X_T range. > > - - caption to Fig. 1: " neutral meson background with charged tracks .... > without charged" may be cryptic to someone not familiar with the analysis; > neutral mesons do not leave tracks after all... The 1800 GeV paper used > "jets with/without..." > Changed to 1800 terminology. > - - p5, par 1: "enhances" -> exploits, emphasizes? > Changed to emphasizes. > - - p5, par 2: "... overlaid with minimum bias..."? The 1800 GeV PRL states > "overlaid with data acquired using a random trigger to model noise, > pileup, underlying event, and multiple pbarp interactions" -- > this implies zero-bias rather than min bias trigger, and gives a more > expanded list of effects being taken care of. BTW, was underlying > event switched off in Pythia, or are we slightly double counting? > (or the UE part is not stated correctly?) Changed to 1800 phrase. > Next sentence: "... neutral background events..." is our jargon, again. > The events were not likely truly neutral; you could move here the > phrase used 2 paragraphs later: "background with and without charged > tracks in the road pointing back to the interaction vertex." > Changed to "background events with and without charged tracks..." > Last sentence: "... match the data....from the photon data." does not > read well. Also, without explanation, may invite questions (in readers > mind) how we use low-purity photon DATA for this purpose. Discarded mention of photon data. > - - p5, last full par: you list components of uncertainties here, and again > on p.6; is this needed? Also, you don't mention comparisons > to theory in Fig. 3. Dropped the second list. The comparison to theory is in the caption of Figure 3, but not in the text. > - - p.6, par 1: "... where E_T^max is the maximum photon transverse energy > in the event." The 1800 GeV paper states "E_T^max is the larger of the > transverse energies of the photon or the leading jet." Which is right? Changed to state what was actually done by Owens and Vogelsang. (The 1800 paper had the above phrase, and it is actually incorrect). > " ... data was FIT to the theoretical cross section" -- I would prefer > FIT-> compared I totally agree. Changed to compared. > "probability that the data fits the theory" -> "probability that the > theory fits the data"? Changed to "theory describes the data". > - - p.6, par 2: expression for dimesionless cross section is different than > previously, could use a predefined symbol. I define a symbol here (\sigma_D) > "The ratio of the dimensionless cross sections (roots=630/roots=1800)" > is likely understandable, but it is NOT the ratio of roots. > If you decide to introduce a symbol for the lengthy formula, you > could write more correctly $\Sigma(roots=630)/\Sigma(roots=1800)" > (yes, picky,picky) I've changed this to use the symbol. > - - p.6, par 3: duplicate list of contributions to uncertainties. > Could be more useful to comment on the level of cancellations > between various 630 and 1800 GeV systematics. Deleted duplicate list. > - - p.7, caption to Fig:5: long formula, and "roots=630/roots=1800" > comments as before. Add: vs x_T? Changed to use symbol. > Fig. 5, Vertical label "630/1800 Scaled" -> Dimensionless > (or use symbols) Now use symbols > - - p.7, par 1: "the data was fit to the theoretical" -> "the theoretical > ... was fit to the data". Again "... probability that theory fits the data" Changed to "comparison of theoretical cross section to data...". > "the overall results" -> "at higher x_T the results" > In the 1800 GeV paper we decided not to brush aside the possibly > interesting disagreement between NLO QCD and data at low pt! I added one short phrase, though the low x_T discrepency is pretty minor in this paper. > - - p.7, par 2, last sentence: "... and the theoretical prediction at higher > E_T values" (same reason). Same as previous comment. > Hope this helps. > Keep us informed about new versions/plots/developments. > Cheers, > Marek > > -Mike --------------------------------------------------------------------------