Some Globe Instructions Making opt tree 1) Define your output tree branches in AnalysisScripts/common/treevariables.dat Example line: type=0 name=cur_type where type=0,1,2,3 (int,unsigned int,float,double) http://cmssw.cvs.cern.ch/cgi-bin/cmssw.cgi/UserCode/HiggsAnalysis/HiggsTo2photons/h2gglobe/AnalysisScripts/common/treevariables.dat?revision=1.6&view=markup 2) In .cc code, do l.FillTree("name", value); wherever you want to fill the tree. 3) Run globe I recommend the batch que. In AnalysisScripts your three scripts will be: mk_fitter.py, submit_fitter.py and check_fitter.py. ./mk_fitter.py -i pathto/datafiles.dat -n 50 -l mylabel -o mylabel/scriptname ./submit_fitter.py -d mylabel -q 1nd ./check_fitter.py mylabel Also, you can do bjobs to check if your jobs are running. 4) hadd your histogram_mylabel_jobnum.root output that will be in eos. All of your trees will be in here. 5) use merger.C to merge the trees in your hadd'ed file (each sample has a tree) http://cmssw.cvs.cern.ch/cgi-bin/cmssw.cgi/UserCode/HiggsAnalysis/HiggsTo2photons/h2gglobe/UCSDplotter/merger.C?revision=1.1&view=markup Histograms from globe 1) Run globe Same as step 3 in tree making 2) hadd your histogram_mylabel_jobnum.root output that will be in eos. 3) plotvariables and inputfiles' trees will not be in the correct format from the split and hadd'ed files. You will run globe again, with the flag makeDummyTrees=1 in your datafiles.dat. It should be on line alone after the InputBranches lines. Do this interactively. It is fast. python fitter.py -i pathto/mydatafiles.dat 4) histograms_mylocaldummytree.root has the correct format trees. Use copyTreesForPI.C to copy these trees into the hadd'ed root file root .L copyTreesForPI.C++ copyTreesForPI("histograms_hadded.root","histograms_mylocaldummytree.root") 5) Use plotinteractive.py for viewing histograms in histograms_hadded.root Making Workspaces from globe directly 1) Run globe 2) python combiner.py -i mylabel/filestocombine.dat --> The output will be called CMS-HGG.root by default