TWiki> Main Web>TWikiUsers>LucieGAUTHIER>SusyTriggers2012 (2012-06-20, LucieGAUTHIER) EditAttachPDF

Susy triggers

Aim of this page : entry point to documentation relevant to Susy triggers.

News : ...

Work in progress, under construction Please feel free to contact me (lucie.gauthier@cernNOSPAMPLEASE.ch) about any of the points below, these are tools I use. Questions/feedback will help making this page more useful.

contents

Introduction

A trigger study for a given path should address :

  • purpose & strategy (which physics ?)
  • rates, efficiencies, expected purities / fake rate, timing
  • L1 seed(s),
  • control & monitoring triggers
  • backup path ( about half the rate of the main )
  • confDB implementations

Tools

Rates

OpenHlt

  • RateEff macro : to calculate rates using OpenHLT ntuples (flat trees) : RateEff howto
    • quick guide 
cvs co -r  HLTrigger/HLTanalyzers/test   <----usually the latest tag / HEAD is fine. The release does not really matter - it runs ok with CMSSW_5_2_0_pre1
cd HLTrigger/HLTanalyzers/test/RateEff/
cmsenv
tar xvzf libconfig-1.3.tar.gz
cd libconfig-1.3/
./configure
make
cd ../
make
source setup.(c)sh
./OHltRateEff myMenu_cfg.py     <---- to get rates. See examples cfg under HLTrigger/HLTanalyzers/test/cfgs/2011cfgs/
how to : get the lumiScaleFactor used in the cfg ?
  • this parameter is the ratio between the luminosity you want to extrapolate to and the luminosity during the run you use.
  • when you run over OpenHLT ntuples, you can specify a block of lumisections to read : 
runLumiblockList = ( 
    (180250, 240,  500 ) # (runnr, minLumiBlock, maxLumiBlock)
 );
  • then you may get the corresponding luminosity using lumiCalc 
python lumiCalc2.py -i json.txt, where json.txt looks like 
{"180250":[[240,500]]}
this gives you the recorded integrated luminosity, say, L, and you get the lumiScaleFactor to a higher lumi HL (in cm-2 s-1) via :
lumiScaleFactor = HL / (L / (23.3 * number of selected lumisections) )
  • Note : how to choose which lumisections to use ? In general, we take a range of LS over which the prescale column remains the same. You may get the prescale column change from wbm (chose a run via RunTimeSummary and click on PrescaleChange)

L1 rates with nanoDST 

cvs co UserCode/L1TriggerDPG
cd UserCode/L1TriggerDPG/macros
root [0] .x initL1Analysis.C 
root [1] .L L1CrossTriggers.C+  or  .L L1CrossTriggersFromSums.C+    <<<<<<<-------- get these from /afs/cern.ch/user/l/lucieg/public/forSusyTriggers/
root [2] RunL1()
     Work in progress, under constructionTo start playing with the macro and add your own path, look for "TO DO" .
  • for 2012 studies : nanoDST trees
  • Rk : muons are trickier because one has to subtract cosmics contribution...no recipe yet though

Implementing a path in ConfDB - testing paths

  • starting confDB
    • try the first link on 1st link
    • try the second set of instructions on the same page 2nd link
    • command-line style : 
On your local PC in antarctica:
setenv CVSROOT @cmscvs.cern.ch:/cvs_server/repositories/CMSSW
setenv CVS_RSH ssh
cvs co -r V01-06-00 EventFilter/ConfigDB (latest tag : see [[hn][https://hypernews.cern.ch/HyperNews/CMS/get/hlt/2665.html]]
cd EventFilter/ConfigDB
ant (only once)
./startGUI&

or

From lxplus (under $CMSSW_BASE/src for instance)
cvs co -r V01-06-00 EventFilter/ConfigDB (latest tag : see [[hn][https://hypernews.cern.ch/HyperNews/CMS/get/hlt/2665.html]]
cd EventFilter/ConfigDB
ant  (only the first time)
./startGUI&

  • implementing a path :
  • get the config to test the code : SWGuideGlobalHLT
  • Quick guide : running the 5e33 menu
    • python dumps :....things change too fast to keep them up to date...sorry. Better open a new configuration in confDB, import the paths that you're interested in from the latest master table (/dev/CMSSW_5_1_0/GRun/V...) 

follow the instructions "preparing a working area" from the SWGuideGlobalHLT link
edmConfigFromDB --cff --configName /dev/CMSSW_5_1_0/GRun --nopaths > setup_cff.py
hltGetConfiguration /your/test/menu/with/only/paths --full --offline --data --unprescale --process TEST --globaltag auto:hltonline --l1-emulator  --l1Xml L1Menu_Collisions2012_v0_L1T_Scales_20101224_Imp0_0x1027.xml > hlt.py
cmsRun hlt.py

  • Comments :
    • the HLT needs RAW format to run, and can run on RAW+RECO, RAW-RECO, RAW+AOD... as long as there is RAW.
    • the current table is in CMSSW_5_2_0...not guaranteed it's going to run on a sample made with 44 releases.

TimingTools

Samples

edm-like ntuples

OpenHLT ntuples

  • all samples can usually be found under
    • nsls /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/
    • ls /pnfs/cms/WAX/11/store/user/lpctrig/Commish2011/
    • list of samples : (TMD list).
NB : not all of these data ntuples have reco information - check for the name of the directories/files or in the comments on the twiki.

  • latest data :
    • fnal : /pnfs/cms/WAX//11/store/user/lpctrig/Commish2011/r180250__*_Run2011B-PromptReco-v1__20111129_1403/ (or 1404 for some samples)
    • cern : /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/r180250__*_Run2011B-PromptReco-v1__20111129_1403/ (or 1404 for some samples)

  • MC :
    • pile-up 23 (analyzer tag V03-10-03, 442_patch7 : L1 jets, fastjet rho included)
      • LM6 : /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/LM6_SUSY_sftsht_7TeV-pythia6_Fall11-E7TeV_Ave23_50ns-v1 (& fnal)
      • LM9 : /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/LM9_SUSY_sftsht_7TeV-pythia6_Fall11-E7TeV_Ave23_50ns-v1 (& fnal)
      • TTbar : /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/TTJets_TuneZ2_7TeV-pythia6_Fall11-E7TeV_Ave23_50ns-v1/ (& fnal)

    • pile-up 32 (analyzer tag V03-10-03, 442_patch7 : L1 jets, fastjet rho included)
      • LM6 : pending - waiting for transfer to T2
      • LM9 : pending - waiting for transfer to T2
      • TTbar : /castor/cern.ch/user/l/lucieg/OpenHLT/Commish2011/TTJets_TuneZ2_7TeV-pythia6_Fall11-E7TeV_Ave32_50ns-v2_0105/ (& fnal)

nanoDST trees :

    • /castor/cern.ch/user/l/lucieg/L1/L1TreeL1Accept_178208_LS_325_345.root - don't use for muons. Missing BPTX requirement = muon seeds prefiring.
    • /castor/cern.ch/user/l/lucieg/L1/L1TreeL1Accept_179828_LS_374_394.root

Miscellaneous links

wbm : run by run information

luminosity, pile-up, HLT menu, L1 mode, prescale column changes, online rates, daily fits of paths cross-sections... see tutorial

ConfigBrowser :

HLT menu, prescale columns. Menus used online can be found under the ORCOFF tab, following the path given on the run table in wbm (e.g. /cdaq/physics/Run2011/5e33/HLT : click on the arrow to access different version)

correlations @L1

eta-phi definitions at L1

Meetings / talks

2012 meetings should take place on Wednesdays, B weeks, from 2pm to 4pm....but weekly meetings in January to prepare for the 2nd round of trigger reviews (February 9-10th)

First look at 2012 menu : preliminary rates on 8TeV data

Towards 2012 menus

  • preliminary L1 menu : draft0

2011 Menus

-- LucieGAUTHIER - 12-Sep-2011

Attachments Attachments
Topic attachments
I Attachment History Action Size Date Who Comment
PDFpdf rates190595To5e33.pdf r3 r2 r1 manage 121.7 K 2012-04-07 - 13:34 LucieGAUTHIER rates from 190595 extrapolated to 5e33
PDFpdf rates190782To5e33.pdf r1 manage 166.3 K 2012-04-11 - 17:09 LucieGAUTHIER rates from 190782 extrapolated to 5e33
PDFpdf susy84bTo5e33.pdf r1 manage 36.9 K 2012-04-06 - 21:29 LucieGAUTHIER preliminary rates on 8TeV data
Edit | Attach | Watch | Print version | History: r33 < r32 < r31 < r30 < r29 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r33 - 2012-06-20 - LucieGAUTHIER
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2024 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback