kappaanalysis / kappatools Goto Github PK

Hi,

since the makefiles do not specify any dependencies on other targets within KappaTools, they fail with -j>2. But the structure of the makefiles does not allow to just add this dependency to the targe. At least i do not understand it :(

g++ -m64 -O2 -L. -L/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/lib -Wl,-rpath,/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/lib -L/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/external/lib -Wl,-rpath,/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/external/lib -lboost_regex -L/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/external/boost/lib/ -L/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/Kappa/lib -Wl,-rpath,/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/Kappa/lib -L/cvmfs/cms.cern.ch/slc6_amd64_gcc481/lcg/root/5.34.18-cms12/lib -lCore -lCint -lRIO -lNet -lHist -lGraf -lGraf3d -lGpad -lTree -lRint -lPostscript -lMatrix -lPhysics -lMathCore -lThread -pthread -lm -ldl -rdynamic -lGenVector -lReflex -lMathCore -lReflexDict -lKPlotTools -o tests1.sbin tests1.o
/nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/lib/libKRootTools.a /nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/../Kappa/DataFormats/test/libKappa.o /nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/src/KappaTools/lib/libKToolbox.a
/cvmfs/cms.cern.ch/slc6_amd64_gcc481/external/gcc/4.8.1/bin/../lib/gcc/x86_64-redhat-linux-gnu/4.8.1/../../../../x86_64-redhat-linux-gnu/bin/ld: cannot find -lKPlotTools
collect2: error: ld returned 1 exit status
make[1]: *** [tests1.sbin] Error 1
make[1]: *** Waiting for unfinished jobs....
rm tests.o tests1.o

Dear KappaTools experts,

For testing purposes I added a cout line to KappaTools/RootTools/JECTools.h and ran "make -C KappaTools". Nothing happened. I needed to run make with the option -B which is then extremely slow. I think, this needs a fix from somebody, who understands the Makefiles. Thanks.

Cheers,
Thomas

Enable running over a skim which contains empty files

I am pulling in boost using the CMSSW environment. But KappaTools still used its own boost library leading to warnings like this in my code.

/cvmfs/cms.cern.ch/slc6_amd64_gcc481/external/gcc/4.8.1/bin/../lib/gcc/x86_64-redhat-linux-gnu/4.8.1/../../../../x86_64-redhat-linux-gnu/bin/ld: warning: libboost_regex.so.1.57.0, needed by /nfs/dust/cms/user/gsieber/dijetana/ana/CMSSW_7_2_3/external/slc6_amd64_gcc481/lib/libKToolbox.so, may conflict with libboost_regex.so.1.51.0.

Can't we just remove this boost downloading mechanism in KappaTools and require boost to be installed by the user?

Hi,

this concerns:
https://github.com/KappaAnalysis/KappaTools/blob/master/Toolbox/scripts/ini_KappaTools.sh#L35
I tried a fresh checkout. When our ini_Higgs... script calls the ini_KappaTools.sh I get the error:

Setting up kappa version 1.0: Updating symlinks under external/slc6_amd64_gcc481 /afs/desy.de/user/r/rfriese/xxl/analysis71/CMSSW_7_1_5/src/Artus/../KappaTools/Toolbox/scripts/ini_KappaTools.sh: line 35: symlinks: command not found

I never did anything in KappaTools and what this line is supposed to do, if it's linked to anything our checkout script or anything else. The error doesn't seem severe, I can still compile and run Artus.

As Fabio discovered, there is a bug in the FileInterface2, when there are >1 entries in the lumiTree for the same lumisection breaking the matching between the lumiInfo and the corresponding events.

This happens for example when skimming a fixed number of events per job and splitting within a lumisection. This should at least lead to an fatal error or ,even better, get fixed in the FileInterface.