def mva_ratioplot(ndata, nsimu, display):
import sys
hi = PlotObject.DisplayOnScreen(display)
var = 'mva'
fdata = ROOT.TFile.Open(ndata)
fsimu = ROOT.TFile.Open(nsimu)
tdata = fdata.Get('t')
tsimu = fsimu.Get('t')
minval, maxval = PlotObject.VarOptimizedRange(var, tdata)
canv = ROOT.TCanvas('c1', '', 1200, 1000)
canv.cd()
PlotObject.Transparent(canv)
'''
upperpad=PlotObject.UpperPad()
lowerpad=PlotObject.LowerPad()
'''
for region in ['barrel', 'endcap']:
canv.Clear()
hists = histMgr(region)
binning = 40
hists.Create1D('data', binning, minval, maxval)
hists.Create1D('orig', binning, minval, maxval)
hists.Create1D('calb', binning, minval, maxval)
tdata.Draw(
'mva >> %s' % (hists.FullName('data')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'mva_nocorr>> %s' % (hists.FullName('orig')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'mva >> %s' % (hists.FullName('calb')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
canv.Clear()
scaling = lambda hist: hist.Scale(
float(hists['data'].GetEntries()) / float(hist.GetEntries()))
#scaling = lambda hist : hist.Scale( float(hists['data'].Integral(binning/10,binning+1)) / float(hist.Integral(binning/10, binning+1)) )
scaling(hists['orig'])
scaling(hists['calb'])
PlotObject.HistSetting_Visualization_data(hists['data'],
MarkerSize_=1.3)
PlotObject.HistSetting_Visualization(hists['orig'],
LineColor_=44,
LineWidth_=3)
PlotObject.HistSetting_Visualization(hists['calb'],
LineColor_=38,
LineWidth_=3)
ratiohists = histMgr('ratio', region)
ratiohists['orig'] = PlotObject.RatioPlot(hists['data'], hists['orig'])
ratiohists['calb'] = PlotObject.RatioPlot(hists['data'], hists['calb'])
leg = _updownplots(hists, ratiohists)
canv.Update()
canv.SaveAs('ratioplot_%s_%s.pdf' % (region, var))
def myratioplot(ndata, nsimu, display):
import sys
hi = PlotObject.DisplayOnScreen(display)
var = sys.argv[1]
fdata = ROOT.TFile.Open(ndata)
fsimu = ROOT.TFile.Open(nsimu)
tdata = fdata.Get('t')
tsimu = fsimu.Get('t')
minval, maxval = PlotObject.VarOptimizedRange(var, tdata)
canv = ROOT.TCanvas('c1', '', 1200, 1000)
canv.cd()
PlotObject.Transparent(canv)
'''
upperpad=PlotObject.UpperPad()
lowerpad=PlotObject.LowerPad()
'''
for region in ['barrel', 'endcap']:
hists = histMgr(region + '_')
binning = 40
hists.Create1D('data', binning, minval, maxval)
hists.Create1D('orig', binning, minval, maxval)
hists.Create1D('calb', binning, minval, maxval)
tdata.Draw(
'%s >> %s' % (var, hists.FullName('data')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'%s >> %s' % (var, hists.FullName('orig')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'calib_%s>> %s' % (var, hists.FullName('calb')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
scaling = lambda hist: hist.Scale(
float(hists['data'].GetEntries()) / float(hist.GetEntries()))
scaling(hists['orig'])
scaling(hists['calb'])
hists['ratio_orig'] = PlotObject.RatioPlot(hists['data'],
hists['orig'],
xtitle=var,
ytitle='data/MC')
hists['ratio_calb'] = PlotObject.RatioPlot(hists['data'],
hists['calb'],
xtitle=var,
ytitle='data/MC')
PlotObject.HistSetting_data(hists['data'], MarkerSize_=1.3)
PlotObject.HistSetting(hists['orig'], LineColor_=44, LineWidth_=3)
PlotObject.HistSetting(hists['calb'], LineColor_=38, LineWidth_=3)
PlotObject.HistSetting(hists['ratio_orig'],
LineColor_=44,
LineWidth_=3)
PlotObject.HistSetting(hists['ratio_calb'],
LineColor_=38,
LineWidth_=3)
canv.Clear()
upperpad = PlotObject.UpperPad()
lowerpad = PlotObject.LowerPad()
upperpad.Draw()
lowerpad.Draw()
upperpad.cd()
fy_max = hists['data'].GetMaximum()
fx_min = hists['data'].GetBinLowEdge(1)
fx_max = hists['data'].GetBinLowEdge(hists['data'].GetNbinsX() + 1)
fx_range = fx_max - fx_min
hists['data'].SetMaximum(fy_max * 1.5)
hists['data'].SetMinimum(0.)
hists['data'].Draw('axis')
hists['calb'].Draw('same hist')
hists['orig'].Draw('same hist')
hists['data'].Draw('same e0 p0')
leg = PlotObject.Legend((fx_min + fx_range * 0.05, fy_max * 1.1),
(fx_max - fx_range * 0.3, fy_max * 1.48),
title='',
useNDC=False)
leg.SetTextAlign(32)
leg.AddEntry(
hists['ratio_orig'], 'origin val, chi2 / nDoF={:6.2f}'.format(
chi2perDoF(hists['data'], hists['orig'])), 'l')
leg.AddEntry(
hists['ratio_calb'], 'calibrated, chi2 / nDoF={:6.2f}'.format(
chi2perDoF(hists['data'], hists['calb'])), 'l')
leg.Draw()
lowerpad.cd()
hists['ratio_calb'].Draw('axis')
hists['ratio_calb'].Draw('same')
hists['ratio_orig'].Draw('same')
canv.Update()
canv.SaveAs('ratioplot_%s_%s.pdf' % (region, var))
def mva_ratioplot(ndata, nsimu, display):
import sys
hi = PlotObject.DisplayOnScreen(display)
var = 'mva'
fdata = ROOT.TFile.Open(ndata)
fsimu = ROOT.TFile.Open(nsimu)
print ndata
tdata = fdata.Get('t')
tsimu = fsimu.Get('t')
minval, maxval = PlotObject.VarOptimizedRange(var, tdata)
canv = ROOT.TCanvas('c1', '', 1200, 1000)
canv.cd()
PlotObject.Transparent(canv)
for region in ['barrel', 'endcap']:
canv.Clear()
hists = histMgr(region)
binning = 40
hists.Create1D('data', binning, minval, maxval)
hists.Create1D('orig', binning, minval, maxval)
hists.Create1D('calb', binning, minval, maxval)
tdata.Draw(
'mva >> %s' % (hists.FullName('data')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'mva_nocorr>> %s' % (hists.FullName('orig')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
tsimu.Draw(
'mva >> %s' % (hists.FullName('calb')), '&&'.join([
Selections.DrawCutStr_ZmassWindow(),
Selections.DrawCutStr_data_PurifyZ(),
Selections.DrawCutStr_EtaRegion(region)
]))
canv.Clear()
#scaling = lambda hist : hist.Scale( float(hists['data'].GetEntries()) / float(hist.GetEntries()) )
#scaling = lambda hist : hist.Scale( float(hists['data'].Integral(binning/10,binning+1)) / float(hist.Integral(binning/10, binning+1)) )
scaling = lambda hist: hist.Scale(
findNormalization_ignoreFirstBins(hists['data']) /
findNormalization_ignoreFirstBins(hist))
scaling(hists['orig'])
scaling(hists['calb'])
PlotObject.HistSetting_Visualization_data(hists['data'],
MarkerSize_=1.3)
PlotObject.HistSetting_Visualization(hists['orig'],
LineColor_=44,
LineWidth_=3)
PlotObject.HistSetting_Visualization(hists['calb'],
LineColor_=38,
LineWidth_=3)
hists['calb'].GetXaxis().SetTitle('BDT output')
hists['orig'].GetXaxis().SetTitle('BDT output')
ratiohists = histMgr('ratio', region)
# asdf class name should move to RatioTH1(numerator, denominator)
ratiohists['orig'] = PlotObject.RatioPlot(hists['data'], hists['orig'])
ratiohists['calb'] = PlotObject.RatioPlot(hists['data'], hists['calb'])
# function name is needed to be DrawRatioPlots
#leg=_updownplots(hists,ratiohists)
leg, line = _updownplots(hists, ratiohists)
canv.Update()
canv.SaveAs('ratioplot_%s_%s.pdf' % (region, var))