(r12) AlpgenRestructuring < Main

Tags: , view all tags
---++ Introduction

   The !AlpgenInterface is undergoing a complete restructuring, according to the [[http://indico.cern.ch/conferenceDisplay.py?confId=35613][new guidelines for GeneratorInterfaces]].
This page documents this restructuring

---++ New Features

---++ Validation
   * What with using the SLHA mass line? Is this doing something bad?
   Not at all up till now. However, from the update notes from Pythia 6.420:
   <verbatim>
PYSLHA: protected some internal PYTHIA particles from read-in of 
    SLHA mass and DECAY tables, in order to avoid incompatibilities with 
    Pythia's hadronization and hadron decay machinery. The SLHA interface is
    intended for read-in of non-SM physics, so we do not consider it a
    strong restriction to protect Pythia's special internal codes and
    all SM particles with masses less than 20 GeV from being
    overwritten by SLHA read-in. People who do wish to alter, say,
    the decay tables of B mesons, should not use the SLHA tables to
    accomplish this, but rather use Pythia's internal updating routine
    PYUPDA.  A modification of the b-quark mass, for example, would
    need to be handled through the PMAS array.
</verbatim>

   * W + jets
      * Basic mother-daughter relations validation.
      * Basic INORAD validation. 
   * Z + jets
      * Basic mother-daughter relations validation.
      * Basic INORAD validation.
         * Why do we get the complete event record seen by UPVETO?
         * I don't know, but Alpgen correctly disregards the six initial particles before passing them to the calorimeter.
         * It also correctly disregards the daughters of INORAD'd particles.
   * ttbar + jets
      * Basic mother daughter relations validation.
      * Basic INORAD validation.
      * Why is it freezing? Why does it stop freezing when I use !FourVectors instead of doing calculations manually.
      * Always bitten by FORTRAN/C-style array differences.

---+++ Issues
   * The PYUPRE subroutine:
<verbatim>
C...PYUPRE
C...Rearranges contents of the HEPEUP commonblock so that
C...mothers precede daughters and daughters of a decay are
C...listed consecutively.
</verbatim>
   In standalone Pythia, a call to UPEVNT is always followed by a call to PYUPRE. In order to mimic that behavior, we do the same in !AlpgenInterface.
   * What is the "phantom event" at the beginning of DBPART?
      * Solved: it was an extra call of DBPART in a wrong place. 
   *  How to make sure the PDF sets are chosen consistently in ALPGEN and PYTHIA?
      * Answer: we don't. See what happened in the 2009 ALPGEN gamma+jets by Sushil.
   * How to make sure the CM energy is consistently kept around?
      * Answer: we don't. There is no way to access the CM energy from within the =JetMatching= class.
   * What ALPGEN uses for the jet clustering is the HEPEVT common block, NOT the HEPEUP.

---++ ALPGEN processes
   * For ALPGEN v2.13
   * Notice that all of those processes can have + jets in the final state. For some processes this is explicitly mentioned.

| *IHRD* | *Prefix* | *Meaning* |
|  %RED%1%ENDCOLOR%  | wqq | W + pair of heavy quarks |
|  %RED%2%ENDCOLOR%  | zqq | Z + pair of heavy quarks |
|  %RED%3%ENDCOLOR%  | wjet | W + jets |
|  %RED%4%ENDCOLOR%  | zjet | Z + jets |
|  %GREEN%5%ENDCOLOR%  | vbjet | Multi vector bosons (W,Z,H,photon) |
|  %BLUE%6%ENDCOLOR%  | 2Q | Pair of heavy quarks |
|  7  | 4Q | Two pairs of heavy quarks |
|  %ORANGE%8%ENDCOLOR%  | QQh | Pair of heavy quarks + Higgs |
|  9  | njet | Pure multijets |
|  %RED%10%ENDCOLOR%  | wcjet | W + single charm |
|  11  | phjet | photon + jets |
|  12  | hjet | Higgs + jets |
|  %PURPLE%13%ENDCOLOR%  | top | Single top |
|  %RED%14%ENDCOLOR%  | wphjet | W + photon + jets |
|  %RED%15%ENDCOLOR%  | wphqq | W + photon + pair of heavy quarks |
|  16  | 2Qph | Pair of heavy quarks + photon |

   * Functions to fix the Event Record:
%RED%void fixEventWZ(lhef::HEPEUP &hepeup);%ENDCOLOR%
%GREEN%void fixEventMultiBoson(lhef::HEPEUP &hepeup);%ENDCOLOR%
%BLUE%void fixEventTTbar(lhef::HEPEUP &hepeup);%ENDCOLOR%
%ORANGE%void fixEventHiggsTTbar(lhef::HEPEUP &hepeup);%ENDCOLOR%
%PURPLE%void fixEventSingleTop(lhef::HEPEUP &hepeup, double mb, int itopprc);%ENDCOLOR%
 
---++ FORTRAN routines

Here we describe the FORTRAN routines responsible for the matching in standalone matching.

---+++ UPINIT routine
   * I/O
   * Call ALSHCD - returns "PYT"
   * I/O
      * Read until "****"
      * Read IHRD
      * Read six masses (mc, mb, mt, mw, mz, mh)
      * Read key-value pairs (PARVAL[NTMP] array)
      * Read until "****"
      * Read xsec, xsec_error, number of unw events, luminosity
   * Call AHSPAR to setup parameters - this sets EBEAM.
      * This sets every parameter except IHRD, the masses and the jet clustering parameters.
   * PBEAM1=DBLE(EBEAM),  PBEAM2=DBLE(EBEAM), IH1=1 --> these are tramp variables to set up HEPRUP later.
   * Call PDFCONVH (what does it do?)
   * Call ALSHCD
   * Prepares INORAD
   * Gets jet matching criteria
   * Sets up the HEPRUP
   * Calls ALSHIN (passes the six masses to PYTHIA)
   * Sets LPRUP (originally it was 661, now it has a meaning).

   * [[%ATTACHURL%/UPINIT.odg][UPINIT.odg]]: UPINIT
 
   The HEPRUP has to be setup as:

   | *Parameter* | *Value* |
   | BEAM |  p,p,energy of CM  |
   | PDFGUP(1,2) |  -1  |
   | IDWTUP |  3  |
   | XSECUP(1) |  AVGWGT  |
   | XERRUP(1) |  ERRWGT  |
   | XMAXUP(1) |  AVGWGT  |

---+++ UPEVNT routine
---+++ UPVETO routine

-- Main.ThiagoTomei - 20 Mar 2009