Ryan's Globe Reference

Globe Links

1. Link to globe board
2. Link to GlobeHzzAnalysis

GlobeSetup _21X

• Check out the Globe Analyzer

cvs co -rTAG -d Analyzers !UserCode/Simon/Analyzers

Remember to switch to CERN cvs.

cvs co RecoEcal/EgammaCoreTools

Edit interface/Limits.h and uncomment #define CMSSW_VERSION_210 Edit src/GlobeEcalClusters.cc and remove &geometry from localCovariance() call.

• Check out the little Globe

Switch back to the UCSD cvs.

cvs co -rTAG -d globe UserCode/Mpieri/globe
cd globe
cp -p ../Analyzers/GlobeAnalyzer/interface/Limits.h branchdef/.
./scripts/gettreedef
./scripts/getnewtclonesarray
./scripts/gettreebranch.pl

Comment out any analysis that are messing up the build.

./script/commentanalsyis NAME

GlobeSetup _16X

Run the following script

#!/bin/bash

user=rwkelley
start=`pwd`
cmssw=1_6_12
globeAn_ver=V00-01-06
globe_ver=GLOBE_053

#setup cvs repo
export CVSROOT=":ext:$user@uaf-2.t2.ucsd.edu:/nfs-1/ucsd-cvs/"
export CVS_RSH="ssh"

#create CMSSW directory
if [ ! -d "CMSSW_$cmssw" ] ; then
    scramv1 project CMSSW CMSSW_$cmssw
fi
cd $start/CMSSW_$cmssw/src/
eval `scramv1 runtime -sh`

if [ ! -d "Analyzers" ] ; then
    cvs co -r$globeAn_ver -d Analyzers UserCode/Simon/Analyzers
    #Remove Pi Zero Discriminator stuff
    temp=Analyzers/GlobeAnalyzer/interface
    sed "s/^\(#define PIZERODISCRIMINATOR\)$/\/\/\1/g" $temp/Limits.h > $temp/tempFile
    mv $temp/tempFile $temp/Limits.h
    switch="-f"

    export CVSROOT=":pserver:anonymous:98passwd@cmscvs.cern.ch:/cvs_server/repositories/CMSSW"
    cvs co -r V00-07-00-02 DataFormats/EgammaCandidates
    cvs co -r V00-04-05-01 DataFormats/EgammaReco
    cvs co -r CMSSW_1_6_8 DataFormats/EgammaReco/interface/BasicCluster.h
    cvs co -r V00-05-12 RecoEcal/EgammaClusterProducers
    cvs co -r V00-01-11 RecoEgamma/Examples
    cvs co -r V00-00-07 -d RecoEgammaBis UserCode/Matteosan/1_6_7/RecoEgammaBis
    mv Analyzers/GlobeAnalyzer/doc/PiZeroDiscriminatorProducer.cc RecoEcal/EgammaClusterProducers/src/
    mv Analyzers/GlobeAnalyzer/doc/PreshowerClusterShapeProducer.cc RecoEcal/EgammaClusterProducers/src/
    export CVSROOT=":ext:$user@uaf-2.t2.ucsd.edu:/nfs-1/ucsd-cvs/"
fi

#build GlobeAnalyzer
cd $start/CMSSW_$cmssw/src
eval `scramv1 runtime -sh`
scramv1 $switch build -j 5

# get mini globe
cd $start/CMSSW_$cmssw/src
if [ ! -d "globe" ] ; then
    cvs co -r$globe_ver -d globe UserCode/Mpieri/globe
    cp -p Analyzers/GlobeAnalyzer/interface/Limits.h globe/branchdef/.
    cd globe
    make clean; make
else
    cd globe
    make
fi

cd $start
echo "----------------------------------------------------"
echo "You should now have a working Globe setup, congrats!"
echo "----------------------------------------------------"

Comment out any analysis that are messing up the build.

./script/commentanalsyis NAME

Description of Files

filenames.dat

The "config" files for globe are found in Hzz/plot_cstep(red)/filenames.dat

typerun=-2
realtyperun=711
fillminuit=0
inputFilesName=plot_cstep/inputfiles.dat
countersName=plot_cstep/counters.dat
cutsName=plot_cstep/cuts.dat
plotVariablesName=plot_cstep/plotvariables.dat

These field determine the behavior of globe and which files to use.

inputfiles.dat

1 intL=100000000.0 rtree=0 schif.root outputFileName.root
typ=1 ind=0 draw=1 Nam=hzz150 Fil=path/h150zz106.root tot=10000 red=9900 lum=1000000000. xsec=0.0056029 kfac=1. scal=1.

• The First line
  1. [int] → # of files to read
  2. intL → integraded luminosity for normalization [pb^-1]
  3. rtree = 0 no output, 1 reduced ROOT file output
  4. [file] → outut reduced ROOT file name
  5. [file] → histogram ROOT file name
• Subsequent lines
  1. typ → type of the sample. Use this to define different behavior for different type of datasets (signal vs. background, gen vs. reco, etc.)
     • Conventions: 0 = data, > 0 = signal MC, < 0 background MC
  2. ind → index of files. Used for sample broken into different files.
  3. draw → draw (1) or not (0) the histogram
  4. Nam → short name for the file (ex: Zee, Hzz, ttbar) for the legends/cuts/counters
  5. Fil → input ROOT file name
  6. tot → total number of events of the sample corresponding to the cross-section read with the analyzer and reduced to the number of red (see next item, only ratio red/tot is used) (maybe only the first line with the same index is used)
  7. red → total number of events in the file (only ratio red/tot is used)
  8. xsec → σ in [pb] of the [tot] events read from the analyzer
  9. kfac → σ k-factor (counters scale with it)
  10. scal scale factor for the histogram display (to show small signals, counters does not scale it)

counters.dat

1
0 ncat=0 pri=1 0 0 0 All_Events

• The First line
  1. [int] → # of counters
• Subsequent lines
  1. [int] → counter sequence number (not used)
  2. ncat → category of counter
  3. pri → 0 = do not print, 1 = print
  4. 3 x [int] → counter # to be used as denominator to compute three different efficiencies (-1 do not compute)
  5. [string] → counter name

plotvariables.dat

2  plot=3 idummy=0
0   htyp=0 plot=-1 ncat=0  3    0   -0.5    2.5     0.  0.  numHs
1   htyp=0 plot=-1 ncat=0  100  0   149.0   151.0   0.  0.  genH_mass

• The First line
  1. [int] → # of histograms
  2. [plot] → default plot command (see plot variable below)
  3. [idummy] → # of histo to start from (1000 is the do all option from PlotHist? () )
• Subsequent lines
  1. [int] → histogram sequence number (if < 0 then gif/eps/C not produced)
  2. htyp → histgram type
     • TH1F = 0
     • TH1F = 1
     • TProfile = 2
  3. plot → plot command (negative: do not plot)
     • -1 = use the default (set in the first line)
     • 0 = integral normalized to cross section
     • 1 = normalized to 1
     • 2 = normalized to xsection
     • 3 = normalized to # entries (no scale factor)
     • 10 = integral normalized to cross section + sig/bg histos (sig type > 0, data 0, bg < 0 )
  4. ncat → histogram categories, if < 0 the # does not correspond to the # of categories, but it is a link to the list of category names at the end of the file.
  5. 2 x [int] → # bins x y (set y = 0 for 1d histos)
  6. 4 x [float] → histogram boundaries: xmin xmax ymin ymax (set y values = 0.0 for 1d histos)
  7. [string] → variable name (to be filled in [AnalysisName]Functions_cc.h)

cuts.dat

Cuts are only parameters (to avoid hard coding)

3
0  cut_red_etrecel           ncat=1   dir=0  fin=0  5.0
1  cut_red_etagenmu          ncat=1   dir=0  fin=0  5.0
2  cut_genet                 ncat=3   dir=0  fin=0  5.0 3.0 10.0

• The First line
  1. [int] → # of cuts
• Subsequent lines
  1. [int] → cut sequence number
  2. [string] → name of cut
  3. ncat → # cut categories (can be used to have an array of cuts)
  4. dir → direction of the cut
     • 0 = variable > cut1
     • 1 = variable < cut1
     • 2 = cut1 < variable < cut2
  5. fin → to use for the (n-1) plots as a final cut (set >=1) or off (0)
  6. [float] → cut1 value
  7. [float] → cut2 value (if needed)