Date: Fri, 29 Sep 2006 10:29:27 +0900 (JST) From: Kaoru Hagiwara To: sg-l__AT__yukawa.kyoto-u.ac.jp, hecforum__AT__jahep.org Subject: [sg-l 3504] HELAS/MadGraph/MadEvent school/mini-WS (2006.10.18-27) Dear colleagues, We are organising a small meeting of physicists who are interested in contributing to LHC physics by making use of the automatic matrix-element/event-generator system HELAS/MadGraph/MadEvent. School/mini-workshop on HELAS/MadGraph/MadEvent 2006.10.18(Wed) -- 2006.10.27(Fri) KEK Theory Division (mainly at Honkan 2F 220) The purpose of the meeting is twofold. 1. To extend the HELAS/MadGraph/MadEvent library to accomodate new physics models with new particles and/or new interactions. 2. To familiarize ourselves with various procedures to match matrix elements and parton shower. Please let us explain the above motivations: 1. When theorists wish to contribute to LHC physics, e.g., by proposing a new model of elementary particles, one immediately faces the difficulty of evaluating its consequences at the LHC, mainly because many quark and gluon initiated subprocesses contribute in proton proton collisions, and also because observable signals require the calculation of the distributions of charged leptons and missing transverse momentum in the decays of new particles, W, Z, and Higgs bosons. This problem can be made manageable with the help of the new HELAS/MadGraph/MadEvent program package. The UCLouvain group has developed a scheme where a user prepares one's own library of new particles and new interactions, which is combined with the MadGraph which already contains all the standard model particles and their standard interactions. This has already been done for the MSSM and general two Higgs doublet models, and there are many groups who are working on different little Higgs models and various models with large extra space dimensions. We will learn how the present HELAS/MadGraph/MadEvent system allows users to introduce ones own model, by using a universal template called `user model (USRMOD)', and explore the possiblity of further improvements. 2. The second purpose is aimed at both experimentalists and theorists, who wish to understand the nature of signals and their backgrounds at the LHC. Because it is of greatest importance to recognize the necessity of calculating multi-jet distributions from exact tree-level matrix elements such that the sum of 1, 2, 3, ... multi-jet cross sections adds up to the inclusive (total) cross section, please let us explain by using a simple example. Typical new physics signals which can be discovered at the LHC have high pT charged leptons, high pT isolated photons, and/or large missing pT, with many high pT hadron jets. For instance, if one is interested in studying top anti-top Higgs production, one looks for bbbbWW final states where at least one W decays leptonically. One ends up with l+6j or 2l+4j events with missing pT from missing neutrinos. One wishes to employ the signal properties, such as m(jj)=mW, m(bjj)=mt, mT(l+missing pT) \sim mW, mT(bl,missing pT) \sim mt, m(bb)=mH, etc, to select signals over backgrounds. The question is how the background would look like. The QCD background events are calculated from W+(6 or more jets), or WW+(4 or more jets). Exact matrix elements are necessary to obtain the right correlations among W and 6 jets, or among 2 W and 4 jets. W+6j or WW+4j events may be calculated by using any of the matrix element generators like MadGraph. The resulting cross section, however, depends strongly on the jet definition, such as the minimum pT and the jet separation variable R, because the tree-level cross section blows up when two partons get collinear or one gluon gets soft. One needs to have a scheme where each W+nj event has the right distribution according to the QCD matrix elements, whereas the sum of W+1j, W+2j, W+3j, ... events agrees with the inclusive W+anything cross section. This should be realized such that the final hadron events do not depend much on the jet definition parameters. A consistent program has been proposed sometime ago by CKKW [Catani, Krauss, Kuhn, Webber, JHEP 0111, 063(2001), Krauss, JHEP 0208, 015(2002)], and it has been implemented by various groups. Since all available algorithms make use of a parton shower generator to `re-weight' the n-jet cross section, they `match matrix elements and parton showers'. The original CKKW scheme has been adopted in the event generator Sherpa, whereas an alternative MLM prescription is used in another generator ALPGEN. We do not need these newly developed event generators, because the matching schemes can be implemented in any combination of a matrix element calculator and a parton shower generator, such as the combination of MadGraph and Pythia. We will have 4 lecturers to cover the above subjects and one lecturer on NLO QCD corrections. Jun-ichi Kanzaki (KEK) will give a general introduction to HELAS ver.3 and its basic structure. Kaoru Hagiwara (KEK) will explain how to make new HELAS subroutines, by using a few examples: charge conjugation subroutines for amplitudes with fermion number violation (Majorana neutrinos, neutralinos and charginos, etc), magnetic/electric dipole couplings, spin 3/2 (gravitino), spin 2 (graviton) wave functions and their interactions. Michel Herquet (Louvain, Belgium) will give a general introduction to MadGraph/MadEvent 4.0, which has just been completed. Its main new features include additional new models on the theory side, and an effective interface to hadronization/detector simulation on the experimental side. He will also explain how to implement new models by preparing a user library of new particles and new interactions. Johan Alwall (SLAC, USA) will give a general introduction to the problem of matching the matrix elements with parton shower, CKKW and its various implementations. We should learn how both the double counting and dead regions are avoided, and also how the final hadronic events do not depend of the jet definitions. Since tons of small details should be fixed in order to produce multi-jet events right, we would like to ask him to teach us all what he learned from matching the MadGraph matrix elements with the Pythia parton shower generator. We need to know those details in order to understand the systematics of background simulation. Please refer to Johan Alwall's talk at DIS2006 (2006.04.20--24, Tsukuba), session HFS-8, available from the DIS2006 website http://www-conf.kek.jp/dis06/ Barbara Jaeger (Karlsruhe, Germany) will explain to us her recent calculation of the NLO correction to the 2-jet plus 4-lepton processes, pp -> jjllll (l=e,\mu,\nu), which contains the vector-boson fusion subprocesses, qq -> qqVV (V=W or Z). We would like to ask her to teach us details of the dipole subtraction scheme of Catani & Seymour, and on some details about the calculation of pentagon contributions. We also wish to learn why her computer code runs 50 times faster than MadGraph, which should tell us how to make a dedicated fast code when it is necessary. We would like to learn how the kinematical distributions among 6 final state partons are affected at the NLO level. Although it is not yet within our scope to implement the parton-shower matching at NLO-QCD accuracy, we would like to have discussions along this line of thoughts. Several groups are starting to work in this direction. Those who wish to attend the school/mini-workshop should contact one of the two organizers: Kaoru Hagiwara: kaoru.hagiwara at kek.jp Junichi Kanzaki: junichi.kanzaki at kek.jp We request all the attendants to have their own PC with wireless LAN card. All the lectures and discussions will be done in English. We encourage all the attendants to be familiarize with MadGraph/MadEvent 4.0 on the website: http://madgraph.phys.ucl.ac.be/ before coming to KEK. The manual for the version 4.0 is not ready, but you may learn from README files. We are now preparing for a new helas/madgraph/madevent website at KEK, and this school/mini- workshop is organized as a part of this project. Unfortunately, we do not have budget to support the participants. Please obtain support from elsewhere. Please send us your willingness to attend as soon as possible, so that we can book your room at the KEK dormitory. Yours sincerely, Kaoru Hagiwara and Junichi Kanzaki %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Date: Fri, 13 Oct 2006 12:52:40 +0900 (JST) From: Kaoru Hagiwara To: sg-l__AT__yukawa.kyoto-u.ac.jp, hecforum__AT__jahep.org Subject: [sg-l 3517] HELAS/MadGraph/MadEvent school/mini-WS (2006.10.18-27) This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. ---1463789130-988174484-1160710613=:17269 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Content-ID: Dear colleagues, We are sending you the second and the last general announcement of the school/mini-workshop on HELAS/MadGraph/MadEvent: 2006.10.18(Wed) -- 2006.10.27(Fri) KEK (mainly at Building 4, 4F, room 414) I copy the first announcement of Sep/29 at the bottom of this e-mail. The only change from it is that we now booked a room 414, 4'th floor of the building 4, throughout the meeting period, excluding Friday afternoons (on Oct/20 and Oct/27) when we should move to room 346, 3'rd floor of the building 4. At this moment we plan to start the school on 2006.10.18(Wed) at 09:00 in the room 414, building 4, KEK and close at 2006.10.27(Fri) at 17:00 in the room 346, building 4, KEK but they may change according to the requests from the attendants. We request all the attendants, including those of you at KEK, to register (just send an e-mail to Hagiwara or Kanzaki). From now on, we will communicate only with the registered participants. Those of you who cannot attend but wish to monitor the school/workshop should also send us an e-mail, and you will receive a carbon copy of all our communication among the participants. Although we do not plan to broadcast any of the lectures, we will put all of them on our website so that they can be shared by the participants and those who wish to monitor our activity. As for the motivation of the school/workshop and the baic plans of the lectures, please read the first announcement of Sep/29 which is copied below. Although we have not yet fixed the program, some more details of the contents of the lectures are now available. They include: Jun-ichi Kanzaki (KEK) will give a general introduction to HELAS3 subroutines. Le us quote: I would like to put emphasis on explaining software structures of the library and giving a tutorial on how to use the subroutines. And, if necessary, I will prepare a simple tutorial on ROOT for novice users. Michel Herquet (Louvain, Belgium) will give a general introduction to MadGraph/MadEvent 4.0, which has just been completed. Let us quote: Lesson 1: MG/ME v4 basics Big picture, Web interface (generate process online, online process run, local run), short description of every important file, ... Tutorial 1: First steps Full online simulation for some SM processes. Lesson 2: MG/ME v4, a closer look MadGraph (how is it working, generated files, ...). MadEvent (how is it working, steps, ...) + compare with others? The Cards system. Pythia, PGS, ERA. Tutorial 2: Playing with plots, Pythia, PGS, ERA, ... Lesson 3: New models implementation What is a MG/ME Model. LH parameter file. Calculators. Using the USRMOD. Tutorial 3: Implementing a new model. Compute some processes in BSM models (typical MSSM, 2HDM, Higgs EFT processes). Understanding model files. Using the USRMOD to implement your own model. Seminar: A real-life example of model implementation, the 2HDM. Johan Alwall (SLAC, USA) will give a general introduction to the problem of matching the matrix elements with parton shower, CKKW and its various implementations. Let us quote: I now have a rather clear idea of what I want to do in my lectures. It will probably be 3 x 90 min lectures, 3 x 90 min hands-on tutorials using MadEvent, Pythia/Herwig and possibly Sherpa and/or AlpGen and maybe an additional "freer" 90 min seminar, discussing different difficulties, like massive quarks, factorisation scale dependence and other stuff that has come up during the lectures and tutorials. We have not answered his inquiry below: In order to make as good as possible a job at the lectures and tutorials, it would be nice to know what I can expect from the participants. On the theoretical side: Can I assume they are familiar with tree-level QCD matrix element calculations? Parton densities? Soft and collinear radiation? Parton showers? On the computational side: Can I assume knowledge of Root/PAW? Fortran? Pythia? Because the needs and backgrounds may be different between theorists and experimentalist attendants, we would like to transfer requests from the participants to the lecturers. Barbara Jaeger (Karlsruhe, Germany) will explain to us her recent calculation of the NLO correction to the 2-jet plus 4-lepton processes, pp -> jjllll (l=e,\mu,\nu), which contains the vector-boson fusion subprocesses, qq -> qqVV (V=W or Z). We asked her to prepare the main body of the lectures for theorist attendants such that we can learn all the technical details of the NLO computation. We may ask her to give a seminar which is aimed at more general audience including experimentalists. Kaoru Hagiwara (KEK) will explain how to make new HELAS subroutines, by using a few examples (some of which are available from HELAS3): charge conjugation subroutines for amplitudes with fermion number violation (Majorana neutrinos, neutralinos and charginos, etc), magnetic/electric dipole couplings, spin 3/2 (gravitino), spin 2 (graviton) wave functions and their interactions. If there are students who are not familiar with FORTRAN, either Hagiwara or Kanzaki will give a brief introduction to FORTRAN by using HELAS subroutines as examples. We will try to prepare additional lectures upon requests from the participants, and we may ask the participants to present their works during the workshop. The lecturers will stay at KEK during the following period: Junichi Kanzaki: Oct/18--Oct/27 (whole period) . Michel Herquet: Oct/18--Oct/26 (leave on Oct/27) . Johan Alwall: Oct/20--Oct/27 (arrive on Oct/19) . Barbara Jaeger: Oct/18--Oct/27 (whole period) . Kaoru Hagiwara: Oct/18--Oct/27 (whole period) . We may schedule some lectures on Saturday, Oct/21, but we will not have any lectures on Sunday, Oct/22. If you have external constraints which do not allow you to attend all the lectures, please do not hesitate to contact us. We will try our best to schedule the lectures to maximize their impacts. Although we do not have budget to support the participants, if you need help in booking a dormitory room at KEK, please let us know. The registration/lodging application forms (both in Japanese and in English) are attached to this e-mail. Please fill out the Japanese forms if you can read/write Japanese letters. Because our secretary will not be available in the next week, please fill them out and send to kaoru.hagiwara__AT__kek.jp. Yours sincerely, Kaoru Hagiwara and Junichi Kanzaki ------------------------------------------------------------------------ The first announcement of the HELAS/MG/ME school/mini-WS at KEK ------------------------------------------------------------------------ Date: Fri, 29 Sep 2006 10:29:27 +0900 (JST) From: Kaoru Hagiwara To: sg-l__AT__yukawa.kyoto-u.ac.jp, hecforum__AT__jahep.org Cc: Junichi Kanzaki Subject: HELAS/MadGraph/MadEvent school/mini-WS (2006.10.18-27) Dear colleagues, We are organising a small meeting of physicists who are interested in contributing to LHC physics by making use of the automatic matrix-element/event-generator system HELAS/MadGraph/MadEvent. School/mini-workshop on HELAS/MadGraph/MadEvent 2006.10.18(Wed) -- 2006.10.27(Fri) KEK Theory Division (mainly at Honkan 2F 220) The purpose of the meeting is twofold. 1. To extend the HELAS/MadGraph/MadEvent library to accomodate new physics models with new particles and/or new interactions. 2. To familiarize ourselves with various procedures to match matrix elements and parton shower. Please let us explain the above motivations: 1. When theorists wish to contribute to LHC physics, e.g., by proposing a new model of elementary particles, one immediately faces the difficulty of evaluating its consequences at the LHC, mainly because many quark and gluon initiated subprocesses contribute in proton proton collisions, and also because observable signals require the calculation of the distributions of charged leptons and missing transverse momentum in the decays of new particles, W, Z, and Higgs bosons. This problem can be made manageable with the help of the new HELAS/MadGraph/MadEvent program package. The UCLouvain group has developed a scheme where a user prepares one's own library of new particles and new interactions, which is combined with the MadGraph which already contains all the standard model particles and their standard interactions. This has already been done for the MSSM and general two Higgs doublet models, and there are many groups who are working on different little Higgs models and various models with large extra space dimensions. We will learn how the present HELAS/MadGraph/MadEvent system allows users to introduce ones own model, by using a universal template called `user model (USRMOD)', and explore the possiblity of further improvements. 2. The second purpose is aimed at both experimentalists and theorists, who wish to understand the nature of signals and their backgrounds at the LHC. Because it is of greatest importance to recognize the necessity of calculating multi-jet distributions from exact tree-level matrix elements such that the sum of 1, 2, 3, ... multi-jet cross sections adds up to the inclusive (total) cross section, please let us explain by using a simple example. Typical new physics signals which can be discovered at the LHC have high pT charged leptons, high pT isolated photons, and/or large missing pT, with many high pT hadron jets. For instance, if one is interested in studying top anti-top Higgs production, one looks for bbbbWW final states where at least one W decays leptonically. One ends up with l+6j or 2l+4j events with missing pT from missing neutrinos. One wishes to employ the signal properties, such as m(jj)=mW, m(bjj)=mt, mT(l+missing pT) \sim mW, mT(bl,missing pT) \sim mt, m(bb)=mH, etc, to select signals over backgrounds. The question is how the background would look like. The QCD background events are calculated from W+(6 or more jets), or WW+(4 or more jets). Exact matrix elements are necessary to obtain the right correlations among W and 6 jets, or among 2 W and 4 jets. W+6j or WW+4j events may be calculated by using any of the matrix element generators like MadGraph. The resulting cross section, however, depends strongly on the jet definition, such as the minimum pT and the jet separation variable R, because the tree-level cross section blows up when two partons get collinear or one gluon gets soft. One needs to have a scheme where each W+nj event has the right distribution according to the QCD matrix elements, whereas the sum of W+1j, W+2j, W+3j, ... events agrees with the inclusive W+anything cross section. This should be realized such that the final hadron events do not depend much on the jet definition parameters. A consistent program has been proposed sometime ago by CKKW [Catani, Krauss, Kuhn, Webber, JHEP 0111, 063(2001), Krauss, JHEP 0208, 015(2002)], and it has been implemented by various groups. Since all available algorithms make use of a parton shower generator to `re-weight' the n-jet cross section, they `match matrix elements and parton showers'. The original CKKW scheme has been adopted in the event generator Sherpa, whereas an alternative MLM prescription is used in another generator ALPGEN. We do not need these newly developed event generators, because the matching schemes can be implemented in any combination of a matrix element calculator and a parton shower generator, such as the combination of MadGraph and Pythia. We will have 4 lecturers to cover the above subjects and one lecturer on NLO QCD corrections. Jun-ichi Kanzaki (KEK) will give a general introduction to HELAS ver.3 and its basic structure. Kaoru Hagiwara (KEK) will explain how to make new HELAS subroutines, by using a few examples: charge conjugation subroutines for amplitudes with fermion number violation (Majorana neutrinos, neutralinos and charginos, etc), magnetic/electric dipole couplings, spin 3/2 (gravitino), spin 2 (graviton) wave functions and their interactions. Michel Herquet (Louvain, Belgium) will give a general introduction to MadGraph/MadEvent 4.0, which has just been completed. Its main new features include additional new models on the theory side, and an effective interface to hadronization/detector simulation on the experimental side. He will also explain how to implement new models by preparing a user library of new particles and new interactions. Johan Alwall (SLAC, USA) will give a general introduction to the problem of matching the matrix elements with parton shower, CKKW and its various implementations. We should learn how both the double counting and dead regions are avoided, and also how the final hadronic events do not depend of the jet definitions. Since tons of small details should be fixed in order to produce multi-jet events right, we would like to ask him to teach us all what he learned from matching the MadGraph matrix elements with the Pythia parton shower generator. We need to know those details in order to understand the systematics of background simulation. Please refer to Johan Alwall's talk at DIS2006 (2006.04.20--24, Tsukuba), session HFS-8, available from the DIS2006 website http://www-conf.kek.jp/dis06/ Barbara Jaeger (Karlsruhe, Germany) will explain to us her recent calculation of the NLO correction to the 2-jet plus 4-lepton processes, pp -> jjllll (l=e,\mu,\nu), which contains the vector-boson fusion subprocesses, qq -> qqVV (V=W or Z). We would like to ask her to teach us details of the dipole subtraction scheme of Catani & Seymour, and on some details about the calculation of pentagon contributions. We also wish to learn why her computer code runs 50 times faster than MadGraph, which should tell us how to make a dedicated fast code when it is necessary. We would like to learn how the kinematical distributions among 6 final state partons are affected at the NLO level. Although it is not yet within our scope to implement the parton-shower matching at NLO-QCD accuracy, we would like to have discussions along this line of thoughts. Several groups are starting to work in this direction. Those who wish to attend the school/mini-workshop should contact one of the two organizers: Kaoru Hagiwara: kaoru.hagiwara at kek.jp Junichi Kanzaki: junichi.kanzaki at kek.jp We request all the attendants to have their own PC with wireless LAN card. All the lectures and discussions will be done in English. We encourage all the attendants to be familiarize with MadGraph/MadEvent 4.0 on the website: http://madgraph.phys.ucl.ac.be/ before coming to KEK. The manual for the version 4.0 is not ready, but you may learn from README files. We are now preparing for a new helas/madgraph/madevent website at KEK, and this school/mini- workshop is organized as a part of this project. Unfortunately, we do not have budget to support the participants. Please obtain support from elsewhere. Please send us your willingness to attend as soon as possible, so that we can book your room at the KEK dormitory. Yours sincerely, Kaoru Hagiwara and Junichi Kanzaki ---1463789130-988174484-1160710613=:17269 Content-Type: TEXT/PLAIN; charset=iso-2022-jp; name=lodging.formj Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: lodging form (Japanese) Content-Disposition: attachment; filename=lodging.formj ＜＜省略しました ---1463789130-988174484-1160710613=:17269 Content-Type: TEXT/PLAIN; charset=us-ascii; name=lodging.form Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: lodging form (English) Content-Disposition: attachment; filename=lodging.form ＜＜省略しました