Merge pull request #14 from brucala/CMGTools-from-CMSSW_7_2_3-crab

tools and integration for CRAB production

CMGTools/TTHAnalysis/BuildFile.xml

@@ -1,14 +1,14 @@
 <use name="CommonTools/Utils"/>
 <use name="TrackingTools/IPTools"/>
 <use name="MagneticField/UniformEngine"/>
+<use name="MagneticField/ParametrizedEngine"/>
 <use name="EgammaAnalysis/ElectronTools"/>
 <use name="SimDataFormats/GeneratorProducts"/>
 <use name="RecoVertex/KalmanVertexFit"/>
 <use name="rootrflx"/>
 <use name="roottmva"/>
 <use name="rootminuit2"/>
 <use name="lhapdf"/>
-<flags LDFLAGS="/afs/cern.ch/cms/slc6_amd64_gcc481/cms/cmssw/CMSSW_7_0_2/lib/slc6_amd64_gcc481/pluginMagneticFieldParametrizedEnginePlugins.so" />
 <export>
   <lib   name="1"/>
 </export>

CMGTools/TTHAnalysis/cfg/crab/FrameworkJobReport.xml

@@ -0,0 +1,17 @@
+<FrameworkJobReport>
+<ReadBranches>
+</ReadBranches>
+<PerformanceReport>
+  <PerformanceSummary Metric="StorageStatistics">
+    <Metric Name="Parameter-untracked-bool-enabled" Value="true"/>
+    <Metric Name="Parameter-untracked-bool-stats" Value="true"/>
+    <Metric Name="Parameter-untracked-string-cacheHint" Value="application-only"/>
+    <Metric Name="Parameter-untracked-string-readHint" Value="auto-detect"/>
+    <Metric Name="ROOT-tfile-read-totalMegabytes" Value="0"/>
+    <Metric Name="ROOT-tfile-write-totalMegabytes" Value="0"/>
+  </PerformanceSummary>
+</PerformanceReport>
+
+<GeneratorInfo>
+</GeneratorInfo>
+</FrameworkJobReport>

CMGTools/TTHAnalysis/cfg/crab/README.txt

@@ -0,0 +1,14 @@
+1) edit ../run_susyMT2.cfg to run over your favorite components
+
+2) edit launchall.py to change the CMG-version/tag and the production name
+
+3) run!!!
+> voms-proxy-init -voms cms --valid=50:00
+Enter GRID pass phrase for this identity: pandolfiniGay
+> python launchall.py
+
+
+Notes: 
+- debugging: choose a single component and set debug=True in heppy_crab_config_env.py
+
+- modify heppy_crab_config.py to run only on your favorite sites

CMGTools/TTHAnalysis/cfg/crab/heppy_config.py

@@ -0,0 +1 @@
+../run_susyMT2_cfg.py

CMGTools/TTHAnalysis/cfg/crab/heppy_crab_config.py

@@ -0,0 +1,30 @@
+from WMCore.Configuration import Configuration
+config = Configuration()
+
+config.section_("General")
+config.General.transferLogs = True
+
+config.section_("JobType")
+config.JobType.pluginName = 'PrivateMC'
+config.JobType.psetName = 'heppy_crab_fake_pset.py'
+config.JobType.scriptExe = 'heppy_crab_script.sh'
+# config.JobType.sendPythonFolder = True  #doesn't work, not supported yet? do it by hand
+import os
+os.system("tar czf python.tar.gz --dereference --directory $CMSSW_BASE python")
+os.system("tar czf cmgdataset.tar.gz --directory $HOME .cmgdataset")
+os.system("tar czf cafpython.tar.gz --directory /afs/cern.ch/cms/caf/ python")
+config.JobType.inputFiles = ['FrameworkJobReport.xml','heppy_config.py','heppy_crab_script.py','cmgdataset.tar.gz', 'python.tar.gz', 'cafpython.tar.gz']
+config.JobType.outputFiles = ['output.log.tgz'] # mt2.root is automatically send because of the pset file
+
+config.section_("Data")
+config.Data.inputDBS = 'global'
+config.Data.splitting = 'EventBased'
+config.Data.outLFN = '/store/user/' + os.environ["USER"]
+config.Data.publication = False
+
+config.section_("Site")
+#config.Site.whitelist = ["T2_CH_CSCS"]
+config.Site.whitelist = ["T2_CH_CSCS", "T2_IT_Legnaro", "T2_UK_London_IC", "T2_UK_SGrid_Bristol", "T2_DE_DESY", "T2_ES_CIEMAT", "T2_IT_Rome", "T2_UK_London_Brunel", "T2_AT_Vienna","T2_DE_RWTH","T2_FR_GRIF_IRFU", "T2_HU_Budapest"] #, "T2_FR_IPHC", "T2_BE_IIHE", "T2_IT_Pisa", "T2_ES_IFCA"] , "T2_PK_NCP", "T2_US_Purdue", "T2_UA_KIPT", "T2_KR_KNU", "T2_RU_IHEP", "T2_RU_INR", "T2_RU_JINR", "T2_US_MIT", "T2_US_Wisconsin", "T2_US_UCSD", "T2_US_Vanderbilt", "T2_US_Caltech"]
+config.Site.storageSite = "T3_CH_PSI"
+#config.Data.ignoreLocality = True
+

CMGTools/TTHAnalysis/cfg/crab/heppy_crab_config_env.py

@@ -0,0 +1,47 @@
+# here we set all crab options that are not fixed
+# values we'll be taken from environment variables set in launchall.py
+# fixed options will be taken from heppy_crab_config.py
+
+debug = False
+
+import imp
+file = open( "heppy_crab_config.py", 'r' )
+cfg = imp.load_source( 'cfg', "heppy_crab_config.py", file)
+config = cfg.config
+import os
+import re
+dataset=os.environ["DATASET"]
+NJOBS=int(os.environ["NJOBS"])
+production_label = os.environ["PROD_LABEL"]
+cmg_version = os.environ["CMG_VERSION"]
+
+print dataset, NJOBS
+
+if debug:
+    NJOBS = 2
+    NEVENTS = 200
+
+config.General.requestName = dataset + "_" + cmg_version # task name
+config.General.workArea = 'crab_' + production_label # crab dir name
+
+# this will divide task in *exactly* NJOBS jobs (for this we need JobType.pluginName = 'PrivateMC' and Data.splitting = 'EventBased')
+config.Data.unitsPerJob = 10
+config.Data.totalUnits = config.Data.unitsPerJob * NJOBS
+
+# arguments to pass to scriptExe. They have to be like "arg=value". 
+config.JobType.scriptArgs = ["dataset="+dataset, "total="+str(NJOBS)]
+
+# output will be .../$outLFN/$PRIMARY_DS/$PUBLISH_NAME/$TIMESTAMP/$COUNTER/$FILENAME
+# https://twiki.cern.ch/twiki/bin/view/CMSPublic/Crab3DataHandling
+config.Data.outLFN += '/babies/' + cmg_version
+config.Data.primaryDataset =  production_label
+config.Data.publishDataName = dataset
+#final output: /store/user/$USER/babies/cmg_version/production_label/dataset/150313_114158/0000/foo.bar
+
+# if NEVENTS variable is set then only nevents will be run
+try: 
+    NEVENTS
+except NameError:
+    pass
+else:
+    config.JobType.scriptArgs += ["nevents="+str(NEVENTS)]

CMGTools/TTHAnalysis/cfg/crab/heppy_crab_fake_pset.py

@@ -0,0 +1,6 @@
+import FWCore.ParameterSet.Config as cms
+process = cms.Process('FAKE')
+process.source = cms.Source("PoolSource", fileNames = cms.untracked.vstring())
+process.maxEvents = cms.untracked.PSet(input = cms.untracked.int32(10))
+process.output = cms.OutputModule("PoolOutputModule", fileName = cms.untracked.string('mt2.root'))
+process.out = cms.EndPath(process.output)

CMGTools/TTHAnalysis/cfg/crab/heppy_crab_script.py

@@ -0,0 +1,101 @@
+#!/usr/bin/env python
+import os
+import PhysicsTools.HeppyCore.framework.config as cfg
+#cfg.Analyzer.nosubdir=True
+
+import ROOT
+from DataFormats.FWLite import *
+import sys
+import re
+#import PSet
+
+ROOT.gSystem.Load("libCMGToolsTTHAnalysis")
+ROOT.gSystem.Load("libFWCoreFWLite.so")
+ROOT.gSystem.Load("libDataFormatsFWLite.so")
+ROOT.gSystem.Load("libPhysicsToolsHeppy")
+ROOT.gSystem.Load("libCMGToolsRootTools")
+ROOT.AutoLibraryLoader.enable()
+
+dataset = ""
+total = 0  # total number of jobs for given dataset, not used at the moment
+nevents = None # this means run all events
+nprint  = 0 # quiet printout, change if you want to print the first nprint events
+
+# arguments of scriptExe
+print "ARGV:",sys.argv
+JobNumber=sys.argv[1] # 1st crab argument is jobID
+job = int(JobNumber)
+# if one wants to include more options to be passed to the crab scriptExe add a corresponding argument below
+# crab only allows arguments of the type 'arg=value'
+for arg in sys.argv[2:]:
+    if arg.split("=")[0] == "dataset":  # this argument is strictly necessary
+        dataset = arg.split("=")[1]
+    elif arg.split("=")[0] == "total":
+        total = int(arg.split("=")[1])
+    elif arg.split("=")[0] == "nevents":
+        nevents = int(arg.split("=")[1])
+        print "selected to run over", nevents, "events"
+
+print "dataset:", dataset
+print "job", job , " out of", total
+
+# fetch config file
+import imp
+handle = open("heppy_config.py", 'r')
+cfo = imp.load_source("heppy_config", "heppy_config.py", handle)
+config = cfo.config
+handle.close()
+
+# pick right component from dataset and file from jobID
+selectedComponents = []
+for comp in config.components:
+    if comp.name == dataset:
+        comp.files = comp.files[job-1: job] # first job number is 1
+        comp.name = comp.name+"_Chunk"+str(job)
+        selectedComponents.append(comp)
+
+# check selectedComponents
+if len(selectedComponents) == 0:
+    print "No selected components found!!"
+    print "   - dataset:", dataset
+    print "   - components:", config.components
+if len(selectedComponents)>1:
+    print "More than one selected component:"
+    cfg.printComps(selectedComponents)
+else:
+    print "Selected component:"
+    print selectedComponents[0]
+    print "files: ", selectedComponents[0].files
+
+# set component to run
+config.components = selectedComponents
+
+# run!!!
+from PhysicsTools.HeppyCore.framework.looper import Looper
+looper = Looper( 'Output', config, nPrint = nprint, nEvents = nevents)
+looper.loop()
+looper.write()
+
+#os.system("ls -lR") # for debugging
+
+# assign the right name
+os.rename("Output/mt2.root", "mt2.root")
+
+# print in crab log file the content of the job log files, so one can see it from 'crab getlog'
+print "-"*25
+print "printing output txt files"
+os.system('for i in Output/*.txt; do echo $i; cat $i; echo "---------"; done')
+
+# pack job log files to be sent to output site
+os.system("tar czf output.log.tgz Output/")
+#os.system("mkdir log")
+#os.rename("output.log.tgz log/output.log.tgz")
+
+
+import ROOT
+f=ROOT.TFile.Open('mt2.root')
+entries=f.Get('tree').GetEntries()
+f.Close()
+
+print entries, "events processed"
+print "job succesful"

CMGTools/TTHAnalysis/cfg/crab/heppy_crab_script.sh

@@ -0,0 +1,32 @@
+# extract exported necessary stuff
+tar xzf cmgdataset.tar.gz --directory $HOME
+tar xzf python.tar.gz --directory $CMSSW_BASE
+tar xzf cafpython.tar.gz --directory $CMSSW_BASE
+
+# uncomment for debuging purposes
+
+#ls -lR .
+#echo "ARGS:"
+#echo $@
+#echo "ENV..................................."
+#env
+#echo "VOMS"
+#voms-proxy-info -all
+#echo "CMSSW BASE, python path, pwd, home"
+#echo $CMSSW_BASE
+#echo $PYTHONPATH
+#echo $PWD
+#echo $HOME
+
+# copy auxiliarity data to the right place (json, pu, lep eff, jet corr, ...)
+cp lib/slc*/* $CMSSW_BASE/lib/slc*
+for i in `find src/ -name data -type d`
+do
+    echo $i
+    mkdir -p  $CMSSW_BASE/$i
+    cp -r $i/* $CMSSW_BASE/$i
+done
+
+#ls -lR 
+
+python heppy_crab_script.py $@

CMGTools/TTHAnalysis/cfg/crab/launchall.py

@@ -0,0 +1,29 @@
+import imp, os
+
+# datasets to run as defined from run_susyMT2.cfg
+# right now configured for maximal job splitting
+# in principle one only needs to modify the following two lines:
+production_label = "fullProd_test3"
+cmg_version = 'MT2_CMGTools-from-CMSSW_7_2_3'
+
+handle = open("heppy_config.py", 'r')
+cfo = imp.load_source("heppy_config", "heppy_config.py", handle)
+conf = cfo.config
+handle.close()
+
+os.system("scramv1 runtime -sh")
+os.system("source /cvmfs/cms.cern.ch/crab3/crab.sh")
+
+os.environ["PROD_LABEL"] = production_label
+os.environ["CMG_VERSION"] = cmg_version
+
+for comp in conf.components:
+    #set maximal splitting
+    NJOBS = len(comp.files)
+    os.environ["NJOBS"] = str(NJOBS)
+    os.environ["DATASET"] = str(comp.name)
+    os.system("crab submit -c heppy_crab_config_env.py")
+
+os.system("rm -f python.tar.gz")
+os.system("rm -f cmgdataset.tar.gz")
+os.system("rm -f cafpython.tar.gz")

CMGTools/TTHAnalysis/python/samples/samples_13TeV_PHYS14.py

@@ -361,7 +361,8 @@
 
 #-----------DATA---------------
 
-dataDir = os.environ['CMSSW_BASE']+"/src/CMGTools/TTHAnalysis/data"
+#dataDir = os.environ['CMSSW_BASE']+"/src/CMGTools/TTHAnalysis/data"
+dataDir = "$CMSSW_BASE/src/CMGTools/TTHAnalysis/data"  # use environmental variable, useful for instance to run on CRAB
 #lumi: 12.21+7.27+0.134 = 19.62 /fb @ 8TeV
 
 json=dataDir+'/json/Cert_Run2012ABCD_22Jan2013ReReco.json'

CMGTools/TTHAnalysis/src/SignedImpactParameter.cc

@@ -1,6 +1,6 @@
 #include "CMGTools/TTHAnalysis/interface/SignedImpactParameter.h"
 #include "MagneticField/UniformEngine/src/UniformMagneticField.h"
-#include "MagneticField/ParametrizedEngine/plugins/OAEParametrizedMagneticField.h"
+#include "MagneticField/ParametrizedEngine/src/OAEParametrizedMagneticField.h"
 #include "TrackingTools/PatternTools/interface/TwoTrackMinimumDistance.h"
 #include "TrackingTools/TransientTrack/interface/TransientTrack.h"
 #include "TrackingTools/IPTools/interface/IPTools.h"

MagneticField/ParametrizedEngine/BuildFile.xml

@@ -0,0 +1,9 @@
+<use   name="DataFormats/GeometryVector"/>
+<!--use   name="FWCore/Framework"/-->
+<use   name="FWCore/ParameterSet"/>
+<use   name="MagneticField/Engine"/>
+<use   name="MagneticField/UniformEngine"/>
+<use   name="MagneticField/Records"/>
+<export>
+  <lib   name="1"/>
+</export>

MagneticField/ParametrizedEngine/interface/ParametrizedMagneticFieldFactory.h

@@ -0,0 +1,45 @@
+#ifndef ParametrizedMagneticFieldFactory_h
+#define ParametrizedMagneticFieldFactory_h
+
+/** \class ParametrizedMagneticFieldFactory
+ *
+ *  Create a parametrized field map with the specified configuration.
+ *
+ *  \author N. Amapane - Torino
+ */
+
+
+#include <MagneticField/Engine/interface/MagneticField.h>
+#include <vector>
+#include <string>
+#include <memory>
+
+namespace edm{
+  class ParameterSet;
+}
+
+namespace magneticfield{
+  class ParametrizedMagneticFieldProducer;
+  class VolumeBasedMagneticFieldESProducerFromDB;
+}
+
+class ParametrizedMagneticFieldFactory {
+ public:
+  /// Constructor
+  ParametrizedMagneticFieldFactory();
+
+ private:
+  friend class magneticfield::ParametrizedMagneticFieldProducer;
+  friend class magneticfield::VolumeBasedMagneticFieldESProducerFromDB;
+
+  // Get map configured from pset (deprecated)
+  std::auto_ptr<MagneticField>
+  static get(std::string version, const edm::ParameterSet& parameters);
+
+  // Get map configured from type name and numerical parameters
+  std::auto_ptr<MagneticField>
+  static get(std::string version, std::vector<double> parameters);
+
+};
+#endif
+

MagneticField/ParametrizedEngine/plugins/BuildFile.xml

@@ -3,6 +3,7 @@
 <use   name="FWCore/ParameterSet"/>
 <use   name="MagneticField/Engine"/>
 <use   name="MagneticField/Records"/>
+<use   name="MagneticField/ParametrizedEngine"/>
 <library   file="*.cc" name="MagneticFieldParametrizedEnginePlugins">
   <flags   EDM_PLUGIN="1"/>
 </library>