Version 1 (modified by trac, 8 years ago) (diff)


Development Area

Please use these pages as a "scratch" notebook to report status and eventual problems related to your process generation. If not yet done, please create "your" page by clicking on the interrogation mark. Feel free to organize your page as you want (and to add links, plots, ...), but avoid creating additional pages if not absolutely needed. Please also update your status on this page on a regular basis (started, ongoing generation, first generation finished, game over, ...)

Grid package

Work distribution

  • Michel (michel.herquet@…): cluster support and user support
  • Simon (devisscher@…): supervision on matching parameters choices and validation with Software.MatchChecker

Status summary

Phase 1

In this phase we validate the basic processes:

  • W,Z + jets : Loic & Nicolas
  • VV samples + jets : Pierre & Olivier . Note: here we try to define the samples in a smart way based on the number of charged leptons in the final state.
  • tt+ jets : Julien .
  • H+ jets (HEFT) : Rik and Marco

How to?

There are several points to follow in order to do things fast and safely:

  • The Madevent events have to be produced with the web interface (of course)
  • In order to save place on the data storage of MadGraph, the pythia step will be performed on the fynu storage mass
  • At the adress /nfs/cms/mass11/p/projects/MGProdValidation/test, a directory MG_ME contains everything to perform the pythia step:

The arborescence is MG_ME and then the directory Wplus, Wminus, ttbar and Z. In one of these there is the directory Events where all .lhe have to be stored. The easiest and cleanest way is maybe to put thel in separate dir like 0jet, 1jet...etc...

Knowing this, how to do things? The proposed solution is:

  • Download all LHE files from the web in the corresponding directory and choose a cool name for each. The possibility is for example KindOfProcess_JetMuliplicity_xqcut_NumberOfTheFile. For example: wplus_2jet_10_5.lhe. Of course, the corresponding .hep file (pythia output) should have the samme name, except the extension.
  • In the process directory, there is the perl script that will read the list of files liste.txt containing the adress of all your samples (ex: 2jet/wplus_2jet_10_5.lhe). It produces a shell script: script.dat. This script will perform pythia over each of the lhe file accordind to the pythia card you have choosen
  • The pythia_card.dat has to be put in the process directory and indicated in the perl script
  • perl
  • chmod u+x script.dat
  • ./script.dat

This will execute in a serial way pythia on each sample. This will take a bit of time, maybe a few hour in all a production.

Common parameters to use

  • PDF: CTEQ6L1 -> final PDF to use!
  • fact. and ren. scale: F

Clusters status

Status of the production

  • Friday 8th february

-- Main.OlivierMattelaer - 25 Jan 2008