Пример #1
0
def getMinMax(histosOrGraphs=[]) :
    def mM(obj) :
        cname = obj.Class().GetName()
        if   cname.startswith('TH1') :               return getMinMaxFromTH1(obj)
        elif cname.startswith('TGraphAsymmErrors') : return getMinMaxFromTGraphAsymmErrors(obj)
        elif cname.startswith('TGraph') :            return getMinMaxFromTGraph(obj)
    ms, Ms = verticalSlice([mM(o) for o in histosOrGraphs if o])
    return min(ms), max(Ms)
Пример #2
0
def getMinMax(histosOrGraphs=[]) :
    def mM(obj) :
        cname = obj.Class().GetName()
        if   cname.startswith('TH1') :               return getMinMaxFromTH1(obj)
        elif cname.startswith('TGraphAsymmErrors') : return getMinMaxFromTGraphAsymmErrors(obj)
        elif cname.startswith('TGraph') :            return getMinMaxFromTGraph(obj)
    ms, Ms = verticalSlice([mM(o) for o in histosOrGraphs if o])
    return min(ms), max(Ms)
def main():
    parser = optparse.OptionParser(usage=usage)
    parser.add_option('-i', '--input-dir', default='./out/fakerate')
    parser.add_option('-o', '--output-dir', default='./out/tight_variables_plots', help='dir for plots')
    parser.add_option('-l', '--lepton', default='el', help='either el or mu')
    parser.add_option('-r', '--region', help='one of the regions for which we saved the fake ntuples')
    parser.add_option('-t', '--tag', help='tag used to select the input files (e.g. Apr_04)')
    parser.add_option('-f', '--fill-histos', action='store_true', default=False, help='force fill (default only if needed)')
    parser.add_option('-v', '--verbose', action='store_true', default=False)
    (options, args) = parser.parse_args()
    inputDir  = options.input_dir
    outputDir = options.output_dir
    lepton    = options.lepton
    region    = options.region
    tag       = options.tag
    verbose   = options.verbose
    if not tag : parser.error('tag is a required option')
    if lepton not in ['el', 'mu'] : parser.error("invalid lepton '%s'"%lepton)
    filestems, treenames = utils.verticalSlice(fakeu.tupleStemsAndNames)
    regions = filestems
    assert region in regions,"invalid region '%s', must be one of %s"%(region, str(regions))
    templateInputFilename = "*_%(region)s_tuple_%(tag)s.root" % {'tag':tag, 'region':region}
    templateOutputFilename =  "%(region)s_%(l)s_tight_plots.root" % {'region':region, 'l':lepton}
    treeName = treenames[regions.index(region)]
    outputDir = outputDir+'/'+region+'/'+lepton # split the output in subdirectories, so we don't overwrite things
    mkdirIfNeeded(outputDir)
    outputFileName = os.path.join(outputDir, templateOutputFilename)
    cacheFileName = outputFileName.replace('.root', '_'+region+'_cache.root')
    doFillHistograms = options.fill_histos or not os.path.exists(cacheFileName)
    optionsToPrint = ['inputDir', 'outputDir', 'region', 'tag', 'doFillHistograms']
    if verbose :
        print "working from %s"%os.getcwd()
        print "being called as : %s"%' '.join(os.sys.argv)
        print "options:\n"+'\n'.join(["%s : %s"%(o, eval(o)) for o in optionsToPrint])
    # collect inputs
    if verbose : print 'input files ',os.path.join(inputDir, templateInputFilename)
    tupleFilenames = glob.glob(os.path.join(inputDir, templateInputFilename))
    samples = setSameGroupForAllData(fastSamplesFromFilenames(tupleFilenames, verbose))
    if not samples : samples = [guessSampleFromFilename(f) for f in tupleFilenames] # if the fast guess didn't work, try the slow one
    samplesPerGroup = collections.defaultdict(list)
    filenamesPerGroup = collections.defaultdict(list)
    for s, f in zip(samples, tupleFilenames) :
        samplesPerGroup[s.group].append(s)
        filenamesPerGroup[s.group].append(f)
    vars = ['pt','eta','d0sig','z0SinTheta','etCone','ptCone','etConeCorr','ptConeCorr']
    vars += ['relEtConeStd', 'relPtConeStd', 'relEtConeMod', 'relPtConeMod']
    groups = samplesPerGroup.keys()
    sources = leptonSources
    #fill histos
    if doFillHistograms :
        lepLabel = "(probe %s)"%lepton
        histosPerGroup = bookHistosPerGroup(vars, groups, lepLabel=lepLabel)
        histosPerSource = bookHistosPerSource(vars, sources, lepLabel=lepLabel)
        for group in groups:
            isData = isDataSample(group)
            filenames = filenamesPerGroup[group]
            histosThisGroup = histosPerGroup[group]
            chain = r.TChain(treeName)
            [chain.Add(fn) for fn in filenames]
            print "%s : %d entries"%(group, chain.GetEntries())
            fillHistos(chain, histosThisGroup, histosPerSource, isData, lepton, group, verbose)
        writeHistos(cacheFileName, {'perGroup':histosPerGroup, 'perSource':histosPerSource}, verbose)
    # compute scale factors
    histosPerGroup = fetchHistos(cacheFileName, histoNames(vars, groups), verbose)
    histosPerSource = fetchHistos(cacheFileName, histoNames(vars, sources), verbose)
    plotStackedHistos(histosPerGroup,  outputDir+'/by_group',  region, colors=SampleUtils.colors, verbose=verbose)
    plotStackedHistos(histosPerSource, outputDir+'/by_source', region, colors=fakeu.colorsFillSources(), verbose=verbose)
    plotIsoComparison(histosPerSource, outputDir+'/',          region, lepton, verbose)
def main():
    parser = optparse.OptionParser(usage=usage)
    parser.add_option('-i', '--input-dir', default='./out/fakerate')
    parser.add_option('-o', '--output-dir', default='./out/fake_scale_factor', help='dir for plots')
    parser.add_option('-l', '--lepton', default='el', help='either el or mu')
    parser.add_option('-r', '--region', help='one of the regions for which we saved the fake ntuples')
    parser.add_option('-t', '--tag', help='tag used to select the input files (e.g. Apr_04)')
    parser.add_option('-T', '--tight-def', help='on-the-fly tight def, one of defs in fakeUtils.py: fakeu.lepIsTight_std, etc.')
    parser.add_option('-f', '--fill-histos', action='store_true', default=False, help='force fill (default only if needed)')
    parser.add_option('-v', '--verbose', action='store_true', default=False)
    (options, args) = parser.parse_args()
    inputDir  = options.input_dir
    outputDir = options.output_dir
    lepton    = options.lepton
    region    = options.region
    tag       = options.tag
    verbose   = options.verbose
    if not tag : parser.error('tag is a required option')
    if lepton not in ['el', 'mu'] : parser.error("invalid lepton '%s'"%lepton)
    filestems, treenames = utils.verticalSlice(fakeu.tupleStemsAndNames)
    regions = filestems
    assert region in regions,"invalid region '%s', must be one of %s"%(region, str(regions))

    templateInputFilename = "*_%(region)s_tuple_%(tag)s.root" % {'tag':tag, 'region':region}
    templateOutputFilename =  "%(region)s_%(l)s_scale_histos.root" % {'region':region, 'l':lepton}
    treeName = treenames[regions.index(region)]
    outputDir = outputDir+'/'+region+'/'+lepton # split the output in subdirectories, so we don't overwrite things
    mkdirIfNeeded(outputDir)
    outputFileName = os.path.join(outputDir, templateOutputFilename)
    cacheFileName = outputFileName.replace('.root', '_'+region+'_cache.root')
    doFillHistograms = options.fill_histos or not os.path.exists(cacheFileName)
    onthefly_tight_def = eval(options.tight_def) if options.tight_def else None # eval will take care of aborting on typos
    optionsToPrint = ['inputDir', 'outputDir', 'region', 'tag', 'doFillHistograms', 'onthefly_tight_def']
    if verbose :
        print "working from %s"%os.getcwd()
        print "being called as : %s"%' '.join(os.sys.argv)
        print "options:\n"+'\n'.join(["%s : %s"%(o, eval(o)) for o in optionsToPrint])
    # collect inputs
    if verbose : print 'input files ',os.path.join(inputDir, templateInputFilename)
    tupleFilenames = glob.glob(os.path.join(inputDir, templateInputFilename))
    samples = setSameGroupForAllData(fastSamplesFromFilenames(tupleFilenames, verbose))
    samplesPerGroup = collections.defaultdict(list)
    filenamesPerGroup = collections.defaultdict(list)
    mkdirIfNeeded(outputDir)
    for s, f in zip(samples, tupleFilenames) :
        samplesPerGroup[s.group].append(s)
        filenamesPerGroup[s.group].append(f)
    vars = ['mt0', 'mt1', 'pt0', 'pt1', 'eta1']
    groups = samplesPerGroup.keys()
    #fill histos
    if doFillHistograms :
        start_time = time.clock()
        num_processed_entries = 0
        histosPerGroup = bookHistos(vars, groups, region=region)
        histosPerSource = bookHistosPerSource(vars, leptonSources, region=region)
        histosPerGroupPerSource = bookHistosPerSamplePerSource(vars, groups, leptonSources, region=region)
        for group in groups:
            isData = isDataSample(group)
            filenames = filenamesPerGroup[group]
            if verbose:
                print " --- group : %s ---".format(group)
                print '\n\t'.join(filenames)
            histosThisGroup = histosPerGroup[group]
            histosThisGroupPerSource = dict((v, histosPerGroupPerSource[v][group]) for v in histosPerGroupPerSource.keys())
            chain = r.TChain(treeName)
            [chain.Add(fn) for fn in filenames]
            if verbose: print "%s : %d entries"%(group, chain.GetEntries())
            num_processed_entries += fillHistos(chain, histosThisGroup, histosPerSource,
                                                histosThisGroupPerSource,
                                                lepton, group, region,
                                                onthefly_tight_def=onthefly_tight_def, verbose=verbose)
        writeHistos(cacheFileName, histosPerGroup, histosPerSource, histosPerGroupPerSource, verbose)
        end_time = time.clock()
        delta_time = end_time - start_time
        if verbose:
            print ("processed {0:d} entries ".format(num_processed_entries)
                   +"in "+("{0:d} min ".format(int(delta_time/60)) if delta_time>60 else
                           "{0:.1f} s ".format(delta_time))
                   +"({0:.1f} kHz)".format(num_processed_entries/delta_time))
    # compute scale factors
    histosPerGroup = fetchHistos(cacheFileName, histoNames(vars, groups, region), verbose)
    histosPerSource = fetchHistos(cacheFileName, histoNamesPerSource(vars, leptonSources, region), verbose)
    histosPerSamplePerSource = fetchHistos(cacheFileName, histoNamesPerSamplePerSource(vars, groups, leptonSources, region), verbose)
    plotStackedHistos(histosPerGroup, outputDir+'/by_group', region, verbose)
    plotStackedHistosSources(histosPerSource, outputDir+'/by_source', region, verbose)
    plotPerSourceEff(histosPerVar=histosPerSource, outputDir=outputDir+'/by_source', lepton=lepton, region=region, verbose=verbose)
    for g in groups:
        hps = dict((v, histosPerSamplePerSource[v][g])for v in vars)
        plotPerSourceEff(histosPerVar=hps, outputDir=outputDir, lepton=lepton, region=region, sample=g, verbose=verbose)


    hn_sf_eta = histoname_sf_vs_eta           (lepton)
    hn_sf_pt  = histoname_sf_vs_pt            (lepton)
    hn_da_eta = histoname_data_fake_eff_vs_eta(lepton)
    hn_da_pt  = histoname_data_fake_eff_vs_pt (lepton)
    objs_eta = subtractRealAndComputeScaleFactor(histosPerGroup, 'eta1', hn_sf_eta, hn_da_eta, outputDir, region, verbose)
    objs_pt  = subtractRealAndComputeScaleFactor(histosPerGroup, 'pt1',  hn_sf_pt,  hn_da_pt,  outputDir, region, verbose)
    rootUtils.writeObjectsToFile(outputFileName, dictSum(objs_eta, objs_pt), verbose)
    if verbose : print "saved scale factors to %s" % outputFileName