| 1 |
/****
|
| 2 |
|
| 3 |
Off peak status (RestrictToMassPeak) :
|
| 4 |
|
| 5 |
x Necessary adaptations identified
|
| 6 |
x Started working on necessary adaptations
|
| 7 |
x Necessary adaptations implemented
|
| 8 |
x Necessary adaptations tested
|
| 9 |
|
| 10 |
DONE!
|
| 11 |
|
| 12 |
|
| 13 |
****/
|
| 14 |
#include <iostream>
|
| 15 |
#include <vector>
|
| 16 |
#include <sys/stat.h>
|
| 17 |
#include <fstream>
|
| 18 |
|
| 19 |
#include <TCut.h>
|
| 20 |
#include <TROOT.h>
|
| 21 |
#include <TCanvas.h>
|
| 22 |
#include <TMath.h>
|
| 23 |
#include <TColor.h>
|
| 24 |
#include <TPaveText.h>
|
| 25 |
#include <TRandom.h>
|
| 26 |
#include <TH1.h>
|
| 27 |
#include <TH2.h>
|
| 28 |
#include <TF1.h>
|
| 29 |
#include <TSQLResult.h>
|
| 30 |
#include <TProfile.h>
|
| 31 |
#include <TSystem.h>
|
| 32 |
#include "LimitDroplet.C"
|
| 33 |
|
| 34 |
//#include "TTbar_stuff.C"
|
| 35 |
using namespace std;
|
| 36 |
|
| 37 |
using namespace PlottingSetup;
|
| 38 |
|
| 39 |
|
| 40 |
void rediscover_the_top(string mcjzb, string datajzb) {
|
| 41 |
dout << "Hi! today we are going to (try to) rediscover the top!" << endl;
|
| 42 |
TCanvas *c3 = new TCanvas("c3","c3");
|
| 43 |
c3->SetLogy(1);
|
| 44 |
vector<float> binning;
|
| 45 |
//binning=allsamples.get_optimal_binsize(mcjzb,cutmass&&cutOSSF&&cutnJets,20,50,800);
|
| 46 |
/*
|
| 47 |
binning.push_back(50);
|
| 48 |
binning.push_back(100);
|
| 49 |
binning.push_back(150);
|
| 50 |
binning.push_back(200);
|
| 51 |
binning.push_back(500);
|
| 52 |
|
| 53 |
|
| 54 |
TH1F *dataprediction = allsamples.Draw("dataprediction", "-"+datajzb, binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data, luminosity);
|
| 55 |
TH1F *puresignal = allsamples.Draw("puresignal", datajzb, binning, "JZB [GeV]", "events", cutmass&&cutOSOF&&cutnJets,data, luminosity);
|
| 56 |
// TH1F *puresignal = allsamples.Draw("puresignal", mcjzb, binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,mc, luminosity,allsamples.FindSample("TTJets"));
|
| 57 |
TH1F *observed = allsamples.Draw("observed", datajzb,binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data,luminosity);
|
| 58 |
/*
|
| 59 |
ofstream myfile;
|
| 60 |
TH1F *ratio = (TH1F*)observed->Clone();
|
| 61 |
ratio->Divide(dataprediction);
|
| 62 |
ratio->GetYaxis()->SetTitle("Ratio obs/pred");
|
| 63 |
ratio->Draw();
|
| 64 |
c3->SaveAs("testratio.png");
|
| 65 |
myfile.open ("ShapeFit_log.txt");
|
| 66 |
establish_upper_limits(observed,dataprediction,puresignal,"LM4",myfile);
|
| 67 |
myfile.close();
|
| 68 |
*/
|
| 69 |
|
| 70 |
|
| 71 |
int nbins=100;
|
| 72 |
float low=0;
|
| 73 |
float hi=500;
|
| 74 |
TCanvas *c4 = new TCanvas("c4","c4",900,900);
|
| 75 |
c4->Divide(2,2);
|
| 76 |
c4->cd(1);
|
| 77 |
c4->cd(1)->SetLogy(1);
|
| 78 |
TH1F *datapredictiont = allsamples.Draw("datapredictiont", "-"+datajzb, nbins,low,hi, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data, luminosity);
|
| 79 |
TH1F *datapredictiono = allsamples.Draw("datapredictiono", "-"+datajzb, nbins,low,hi, "JZB [GeV]", "events", cutmass&&cutOSOF&&cutnJets,data, luminosity);
|
| 80 |
datapredictiont->Add(datapredictiono,-1);
|
| 81 |
dout << "Second way of doing this !!!! Analytical shape to the left :-D" << endl;
|
| 82 |
vector<TF1*> functions = do_cb_fit_to_plot(datapredictiont,10);
|
| 83 |
datapredictiont->SetMarkerColor(kRed);
|
| 84 |
datapredictiont->SetLineColor(kRed);
|
| 85 |
datapredictiont->Draw();
|
| 86 |
functions[1]->Draw("same");
|
| 87 |
TText *title1 = write_title("Top Background Prediction (JZB<0, with osof subtr)");
|
| 88 |
title1->Draw();
|
| 89 |
|
| 90 |
c4->cd(2);
|
| 91 |
c4->cd(2)->SetLogy(1);
|
| 92 |
TH1F *observedt = allsamples.Draw("observedt", datajzb, nbins,low,hi, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data, luminosity);
|
| 93 |
observedt->Draw();
|
| 94 |
datapredictiont->Draw("histo,same");
|
| 95 |
functions[1]->Draw("same");
|
| 96 |
TText *title2 = write_title("Observed and predicted background");
|
| 97 |
title2->Draw();
|
| 98 |
|
| 99 |
c4->cd(3);
|
| 100 |
c4->cd(3)->SetLogy(1);
|
| 101 |
// TH1F *ratio = (TH1F*)observedt->Clone();
|
| 102 |
|
| 103 |
TH1F *analytical_background_prediction= new TH1F("analytical_background_prediction","",nbins,low,hi);
|
| 104 |
for(int i=0;i<=nbins;i++) {
|
| 105 |
analytical_background_prediction->SetBinContent(i+1,functions[1]->Eval(((hi-low)/((float)nbins))*(i+0.5)));
|
| 106 |
analytical_background_prediction->SetBinError(i+1,TMath::Sqrt(functions[1]->Eval(((hi-low)/((float)nbins))*(i+0.5))));
|
| 107 |
}
|
| 108 |
analytical_background_prediction->GetYaxis()->SetTitle("JZB [GeV]");
|
| 109 |
analytical_background_prediction->GetYaxis()->CenterTitle();
|
| 110 |
TH1F *analyticaldrawonly = (TH1F*)analytical_background_prediction->Clone();
|
| 111 |
analytical_background_prediction->SetFillColor(TColor::GetColor("#3399FF"));
|
| 112 |
analytical_background_prediction->SetMarkerSize(0);
|
| 113 |
analytical_background_prediction->Draw("e5");
|
| 114 |
analyticaldrawonly->Draw("histo,same");
|
| 115 |
functions[1]->Draw("same");
|
| 116 |
TText *title3 = write_title("Analytical bg pred histo");
|
| 117 |
title3->Draw();
|
| 118 |
|
| 119 |
c4->cd(4);
|
| 120 |
// c4->cd(4)->SetLogy(1);
|
| 121 |
vector<float> ratio_binning;
|
| 122 |
ratio_binning.push_back(0);
|
| 123 |
ratio_binning.push_back(5);
|
| 124 |
ratio_binning.push_back(10);
|
| 125 |
ratio_binning.push_back(20);
|
| 126 |
ratio_binning.push_back(50);
|
| 127 |
// ratio_binning.push_back(60);
|
| 128 |
/*
|
| 129 |
ratio_binning.push_back(51);
|
| 130 |
ratio_binning.push_back(52);
|
| 131 |
ratio_binning.push_back(53);
|
| 132 |
ratio_binning.push_back(54);
|
| 133 |
ratio_binning.push_back(55);
|
| 134 |
ratio_binning.push_back(56);
|
| 135 |
ratio_binning.push_back(57);
|
| 136 |
ratio_binning.push_back(58);
|
| 137 |
ratio_binning.push_back(59);
|
| 138 |
ratio_binning.push_back(60);
|
| 139 |
// ratio_binning.push_back(70);*/
|
| 140 |
// ratio_binning.push_back(80);
|
| 141 |
// ratio_binning.push_back(90);
|
| 142 |
ratio_binning.push_back(80);
|
| 143 |
// ratio_binning.push_back(110);
|
| 144 |
ratio_binning.push_back(500);
|
| 145 |
|
| 146 |
TH1F *observedtb = allsamples.Draw("observedtb", datajzb, ratio_binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data, luminosity);
|
| 147 |
TH1F *datapredictiontb = allsamples.Draw("datapredictiontb", "-"+datajzb, ratio_binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets,data, luminosity);
|
| 148 |
TH1F *datapredictiontbo = allsamples.Draw("datapredictiontbo", "-"+datajzb, ratio_binning, "JZB [GeV]", "events", cutmass&&cutOSOF&&cutnJets,data, luminosity);
|
| 149 |
datapredictiontb->Add(datapredictiontbo,-1);
|
| 150 |
TH1F *analytical_background_predictionb = allsamples.Draw("analytical_background_predictionb",datajzb, ratio_binning, "JZB [GeV]", "events", cutmass&&cutOSSF&&cutnJets&&"mll<2",data, luminosity);
|
| 151 |
for(int i=0;i<=ratio_binning.size();i++) {
|
| 152 |
analytical_background_predictionb->SetBinContent(i+1,functions[1]->Eval(analytical_background_predictionb->GetBinCenter(i)));
|
| 153 |
analytical_background_predictionb->SetBinError(i+1,TMath::Sqrt(functions[1]->Eval(analytical_background_predictionb->GetBinCenter(i))));
|
| 154 |
}
|
| 155 |
|
| 156 |
TH1F *ratio = (TH1F*) observedtb->Clone();
|
| 157 |
ratio->Divide(datapredictiontb);
|
| 158 |
|
| 159 |
for (int i=0;i<=ratio_binning.size();i++) {
|
| 160 |
dout << observedtb->GetBinLowEdge(i+1) << ";"<<observedtb->GetBinContent(i+1) << ";" << datapredictiontb->GetBinContent(i+1) << " --> " << ratio->GetBinContent(i+1) << "+/-" << ratio->GetBinError(i+1) << endl;
|
| 161 |
}
|
| 162 |
|
| 163 |
// ratio->Divide(datapredictiontb);
|
| 164 |
// ratio->Divide(analytical_background_predictionb);
|
| 165 |
// TGraphAsymmErrors *JZBratio= histRatio(observedtb,analytical_background_predictionb,data,ratio_binning);
|
| 166 |
// ratio->Divide(analytical_background_prediction);
|
| 167 |
// ratio->Divide(datapredictiont);
|
| 168 |
// ratio->GetYaxis()->SetTitle("obs/pred");
|
| 169 |
// JZBratio->Draw("AP");
|
| 170 |
ratio->GetYaxis()->SetRangeUser(0,10);
|
| 171 |
ratio->Draw();
|
| 172 |
//analytical_background_predictionb->Draw();
|
| 173 |
// JZBratio->SetTitle("");
|
| 174 |
TText *title4 = write_title("Ratio of observed to predicted");
|
| 175 |
title4->Draw();
|
| 176 |
|
| 177 |
// CompleteSave(c4,"test/ttbar_discovery_dataprediction___analytical_function");
|
| 178 |
CompleteSave(c4,"test/ttbar_discovery_dataprediction__analytical__new_binning_one_huge_bin_from_80");
|
| 179 |
|
| 180 |
|
| 181 |
|
| 182 |
|
| 183 |
|
| 184 |
}
|
| 185 |
|
| 186 |
vector<float> compute_one_upper_limit(float mceff,float mcefferr, int ibin, string mcjzb, string plotfilename, bool doexpected, int flipped) {
|
| 187 |
float sigma95=-9.9,sigma95A=-9.9;
|
| 188 |
/*
|
| 189 |
USAGE OF ROOSTATS_CL95
|
| 190 |
" Double_t limit = roostats_cl95(ilum, slum, eff, seff, bck, sbck, n, gauss = false, nuisanceModel, method, plotFileName, seed); \n"
|
| 191 |
" LimitResult expected_limit = roostats_clm(ilum, slum, eff, seff, bck, sbck, ntoys, nuisanceModel, method, seed); \n"
|
| 192 |
" Double_t average_limit = roostats_cla(ilum, slum, eff, seff, bck, sbck, nuisanceModel, method, seed); \n"
|
| 193 |
" \n"
|
| 194 |
"
|
| 195 |
" Double_t obs_limit = limit.GetObservedLimit(); \n"
|
| 196 |
" Double_t exp_limit = limit.GetExpectedLimit(); \n"
|
| 197 |
" Double_t exp_up = limit.GetOneSigmaHighRange(); \n"
|
| 198 |
" Double_t exp_down = limit.GetOneSigmaLowRange(); \n"
|
| 199 |
" Double_t exp_2up = limit.GetTwoSigmaHighRange(); \n"
|
| 200 |
" Double_t exp_2down = limit.GetTwoSigmaLowRange(); \n"
|
| 201 |
*/
|
| 202 |
if(mceff<=0) {
|
| 203 |
write_warning(__FUNCTION__,"Cannot compute upper limit in this configuration as the efficiency is negative:");
|
| 204 |
dout << "mc efficiency=" << mceff << " +/- " << mcefferr;
|
| 205 |
vector<float> sigmas;
|
| 206 |
sigmas.push_back(-1);
|
| 207 |
sigmas.push_back(-1);
|
| 208 |
return sigmas;
|
| 209 |
} else {
|
| 210 |
int nlimittoysused=1;
|
| 211 |
|
| 212 |
///------------------------------------------ < NEW > ----------------------------------------------------------
|
| 213 |
|
| 214 |
int secondssince1970=time(NULL);
|
| 215 |
stringstream repname;
|
| 216 |
repname << PlottingSetup::cbafbasedir << "/exchange/report_" << secondssince1970 << "_"<<plotfilename<< "__"<< ".txt";
|
| 217 |
|
| 218 |
/* - report filename [1]
|
| 219 |
- luminosity [2]
|
| 220 |
- lumi uncert [3]
|
| 221 |
- MC efficiency [4]
|
| 222 |
- MC efficiency error [5]
|
| 223 |
- Npred [6]
|
| 224 |
- Nprederr [7]
|
| 225 |
- Nobs [8]
|
| 226 |
- JZB cut [9]
|
| 227 |
- plot name [10]*/
|
| 228 |
|
| 229 |
if(flipped==0) dout << "Calling limit capsule instead of calling : CL95(" << luminosity << "," << lumiuncert*luminosity << "," << mceff << "," << mcefferr << "," << Npred[ibin] << "," << Nprederr[ibin] << "," << Nobs[ibin] << "," << false << "," << nuisancemodel<< ") " << endl;
|
| 230 |
if(flipped>0) dout << "Calling limit capsule instead of calling : CL95(" << luminosity << "," << lumiuncert*luminosity << "," << mceff << "," << mcefferr << "," << flippedNpred[ibin] << "," << flippedNprederr[ibin] << "," << flippedNobs[ibin] << "," << false << "," << nuisancemodel<< ") " << endl;
|
| 231 |
|
| 232 |
stringstream command;
|
| 233 |
if(flipped==0) command << PlottingSetup::cbafbasedir << "/DistributedModelCalculations/Limits/TimedLimitCapsule.exec " << repname.str() << " " << luminosity << " " << luminosity*lumiuncert << " " << mceff << " " << mcefferr << " " << Npred[ibin] << " " << Nprederr[ibin] << " " << Nobs[ibin] << " " << -1 << " " << PlottingSetup::basedirectory << "/" << plotfilename << " " << doexpected;
|
| 234 |
if(flipped>0) command << PlottingSetup::cbafbasedir << "/DistributedModelCalculations/Limits/TimedLimitCapsule.exec " << repname.str() << " " << luminosity << " " << luminosity*lumiuncert << " " << mceff << " " << mcefferr << " " << flippedNpred[ibin] << " " << flippedNprederr[ibin] << " " << flippedNobs[ibin] << " " << -1 << " " << PlottingSetup::basedirectory << "/" << plotfilename << " " << doexpected;
|
| 235 |
dout << command.str() << endl;
|
| 236 |
|
| 237 |
int retval = 256;
|
| 238 |
int attempts=0;
|
| 239 |
while(!(retval==0||attempts>=3)) {//try up to 3 times
|
| 240 |
attempts++;
|
| 241 |
dout << "Starting limit calculation (TimedLimitCapsule) now : Attempt " << attempts << endl;
|
| 242 |
retval=gSystem->Exec(command.str().c_str());
|
| 243 |
}
|
| 244 |
char hostname[1023];
|
| 245 |
string completehostname=GetCompleteHostname();
|
| 246 |
gethostname(hostname,1023);
|
| 247 |
if((!((Contains(hostname,"t3ui")||Contains(hostname,"t3wn"))))&&retval==256) {
|
| 248 |
//running via CRAB and encountered the same problem too often: place a problem file to mark this problem!
|
| 249 |
stringstream markproblem;
|
| 250 |
markproblem << "touch " << PlottingSetup::basedirectory << "/exchange/problemswhilesettinglimits.txt";
|
| 251 |
gSystem->Exec(markproblem.str().c_str());
|
| 252 |
}
|
| 253 |
LimitDroplet limres;
|
| 254 |
limres.readDroplet(repname.str());
|
| 255 |
dout << limres << endl;
|
| 256 |
remove(repname.str().c_str());
|
| 257 |
sigma95=limres.observed;
|
| 258 |
|
| 259 |
|
| 260 |
///------------------------------------------ < /NEW > ----------------------------------------------------------
|
| 261 |
vector<float> sigmas;
|
| 262 |
sigmas.push_back(sigma95);
|
| 263 |
if(doexpected) {
|
| 264 |
sigmas.push_back(limres.expected);
|
| 265 |
sigmas.push_back(limres.upper68);
|
| 266 |
sigmas.push_back(limres.lower68);
|
| 267 |
sigmas.push_back(limres.upper95);
|
| 268 |
sigmas.push_back(limres.lower95);
|
| 269 |
}
|
| 270 |
|
| 271 |
return sigmas;
|
| 272 |
|
| 273 |
|
| 274 |
}//end of mc efficiency is ok
|
| 275 |
}
|
| 276 |
|
| 277 |
void compute_upper_limits_from_counting_experiment(vector<vector<float> > uncertainties,vector<float> jzbcuts, string mcjzb, bool doexpected, int flipped) {
|
| 278 |
dout << "Doing counting experiment ... " << endl;
|
| 279 |
vector<vector<string> > limits;
|
| 280 |
vector<vector<float> > vlimits;
|
| 281 |
|
| 282 |
|
| 283 |
for(int isample=0;isample<signalsamples.collection.size();isample++) {
|
| 284 |
vector<string> rows;
|
| 285 |
vector<float> vrows;
|
| 286 |
dout << "Considering sample " << signalsamples.collection[isample].samplename << endl;
|
| 287 |
rows.push_back(signalsamples.collection[isample].samplename);
|
| 288 |
for(int ibin=0;ibin<jzbcuts.size();ibin++) {
|
| 289 |
dout << "_________________________________________________________________________________" << endl;
|
| 290 |
float JZBcutat=uncertainties[isample*jzbcuts.size()+ibin][0];
|
| 291 |
float mceff=uncertainties[isample*jzbcuts.size()+ibin][1];
|
| 292 |
float staterr=uncertainties[isample*jzbcuts.size()+ibin][2];
|
| 293 |
float systerr=uncertainties[isample*jzbcuts.size()+ibin][3];
|
| 294 |
float toterr =uncertainties[isample*jzbcuts.size()+ibin][4];
|
| 295 |
float observed,observederr,null,result;
|
| 296 |
|
| 297 |
// fill_result_histos(observed,observederr, null,null,null,null,null,null,null,mcjzb,JZBcutat,14000,(int)5,result,(signalsamples.FindSample(signalsamples.collection[isample].filename)),signalsamples);
|
| 298 |
// observed-=result;//this is the actual excess we see!
|
| 299 |
// float expected=observed/luminosity;
|
| 300 |
string plotfilename=(string)(TString(signalsamples.collection[isample].samplename)+TString("___JZB_geq_")+TString(any2string(JZBcutat))+TString(".png"));
|
| 301 |
dout << "Sample: " << signalsamples.collection[isample].samplename << ", JZB>"<<JZBcutat<< " : " << mceff << " +/- " << staterr << " (stat) +/- " << systerr << " (syst) --> toterr = " << toterr << endl;
|
| 302 |
vector<float> sigmas = compute_one_upper_limit(mceff,toterr,ibin,mcjzb,plotfilename,doexpected,flipped);
|
| 303 |
|
| 304 |
if(doexpected) {
|
| 305 |
// rows.push_back(any2string(sigmas[0])+";"+any2string(sigmas[1])+";"+"("+any2string(expected)+")");
|
| 306 |
rows.push_back(any2string(sigmas[0])+";"+any2string(sigmas[1])+";"+"("+any2string(signalsamples.collection[isample].xs)+")");
|
| 307 |
vrows.push_back(sigmas[0]);
|
| 308 |
vrows.push_back(sigmas[1]);
|
| 309 |
// vrows.push_back(expected);
|
| 310 |
vrows.push_back(signalsamples.collection[isample].xs);
|
| 311 |
}
|
| 312 |
else {
|
| 313 |
// rows.push_back(any2string(sigmas[0])+"("+any2string(expected)+")");
|
| 314 |
rows.push_back(any2string(sigmas[0]));
|
| 315 |
vrows.push_back(sigmas[0]);
|
| 316 |
vrows.push_back(signalsamples.collection[isample].xs);
|
| 317 |
// vrows.push_back(expected);
|
| 318 |
}
|
| 319 |
}//end of bin loop
|
| 320 |
limits.push_back(rows);
|
| 321 |
vlimits.push_back(vrows);
|
| 322 |
}//end of sample loop
|
| 323 |
dout << endl << endl << endl << "_________________________________________________________________________________________________" << endl << endl;
|
| 324 |
dout << endl << endl << "PAS table 3: (notation: limit [95%CL])" << endl << endl;
|
| 325 |
dout << "\t";
|
| 326 |
for (int irow=0;irow<jzbcuts.size();irow++) {
|
| 327 |
dout << jzbcuts[irow] << "\t";
|
| 328 |
}
|
| 329 |
dout << endl;
|
| 330 |
for(int irow=0;irow<limits.size();irow++) {
|
| 331 |
for(int ientry=0;ientry<limits[irow].size();ientry++) {
|
| 332 |
if (limits[irow][ientry]>0) dout << limits[irow][ientry] << "\t";
|
| 333 |
else dout << " (N/A) \t";
|
| 334 |
}
|
| 335 |
dout << endl;
|
| 336 |
}
|
| 337 |
|
| 338 |
if(!doexpected) {
|
| 339 |
dout << endl << endl << "LIMITS: (Tex)" << endl;
|
| 340 |
tout << "\\begin{table}[hbtp]" << endl;
|
| 341 |
tout << "\\renewcommand{\\arraystretch}{1.3}" << endl;
|
| 342 |
tout << "\\begin{center}" << endl;
|
| 343 |
tout << "\\caption{Observed upper limits on the cross section of different LM benchmark points " << (ConsiderSignalContaminationForLimits?" (accounting for signal contamination)":" (not accounting for signal contamination)") << "}\\label{tab:lmresults}" << endl;
|
| 344 |
tout << "" << endl;
|
| 345 |
tout << "\\begin{tabular}{ | l | ";
|
| 346 |
for (int irow=0;irow<jzbcuts.size();irow++) tout << " l |";
|
| 347 |
tout << "} " << endl << " \\hline " << endl << "& \t ";
|
| 348 |
for (int irow=0;irow<jzbcuts.size();irow++) {
|
| 349 |
tout << "JZB $>$ " << jzbcuts[irow] << " GeV & \t ";
|
| 350 |
}
|
| 351 |
tout << " \\\\ \\hline " << endl;
|
| 352 |
for(int irow=0;irow<limits.size();irow++) {
|
| 353 |
tout << limits[irow][0] << " \t";
|
| 354 |
for(int ientry=0;ientry<jzbcuts.size();ientry++) {
|
| 355 |
if(vlimits[irow][2*ientry]>0) tout << " & " << Round(vlimits[irow][2*ientry],2) << " \t (" << Round(vlimits[irow][2*ientry] / vlimits[irow][2*ientry+1],3)<< "x \\sigma ) \t";
|
| 356 |
else tout << " & ( N / A ) \t";
|
| 357 |
// dout << Round(vlimits[irow][2*ientry],3) << " / " << Round(vlimits[irow][2*ientry+1],3)<< "\t";
|
| 358 |
}
|
| 359 |
tout << " \\\\ \\hline " << endl;
|
| 360 |
}
|
| 361 |
tout << "\\end{tabular}" << endl;
|
| 362 |
tout << " \\end{tabular}"<< endl;
|
| 363 |
tout << "\\end{center}"<< endl;
|
| 364 |
tout << "\\end{table} "<< endl;
|
| 365 |
|
| 366 |
}//do observed
|
| 367 |
|
| 368 |
dout << endl << endl << "Final selection efficiencies with total statistical and systematic errors, and corresponding observed and expected upper limits (UL) on ($\\sigma\\times$ BR $\\times$ acceptance) for the LM4 and LM8 scenarios, in the different regions. The last column contains the predicted ($\\sigma \\times $BR$\\times$ acceptance) at NLO obtained from Monte Carlo simulation." << endl;
|
| 369 |
dout << "Scenario \t Efficiency [%] \t Upper limits [pb] \t \\sigma [pb]" << endl;
|
| 370 |
for(int icut=0;icut<jzbcuts.size();icut++) {
|
| 371 |
dout << "Region with JZB>" << jzbcuts[icut] << (ConsiderSignalContaminationForLimits?" (accounting for signal contamination)":" (not accounting for signal contamination)") << endl;
|
| 372 |
for(int isample=0;isample<signalsamples.collection.size();isample++) {
|
| 373 |
dout << limits[isample][0] << "\t" << Round(100*uncertainties[isample*jzbcuts.size()+icut][1],3) << "+/-" << Round(100*uncertainties[isample*jzbcuts.size()+icut][2],3) << " (stat) +/- " << Round(100*uncertainties[isample*jzbcuts.size()+icut][3],3) << " (syst) \t" << Round((vlimits[isample][2*icut]),3) << "\t" << Round(vlimits[isample][2*icut+1],3) << endl;
|
| 374 |
}
|
| 375 |
dout << endl;
|
| 376 |
}
|
| 377 |
}
|
| 378 |
|
| 379 |
|
| 380 |
|
| 381 |
/********************************************************************** new : Limits using SHAPES ***********************************
|
| 382 |
|
| 383 |
|
| 384 |
SSSSSSSSSSSSSSS hhhhhhh
|
| 385 |
SS:::::::::::::::Sh:::::h
|
| 386 |
S:::::SSSSSS::::::Sh:::::h
|
| 387 |
S:::::S SSSSSSSh:::::h
|
| 388 |
S:::::S h::::h hhhhh aaaaaaaaaaaaa ppppp ppppppppp eeeeeeeeeeee ssssssssss
|
| 389 |
S:::::S h::::hh:::::hhh a::::::::::::a p::::ppp:::::::::p ee::::::::::::ee ss::::::::::s
|
| 390 |
S::::SSSS h::::::::::::::hh aaaaaaaaa:::::ap:::::::::::::::::p e::::::eeeee:::::eess:::::::::::::s
|
| 391 |
SS::::::SSSSS h:::::::hhh::::::h a::::app::::::ppppp::::::pe::::::e e:::::es::::::ssss:::::s
|
| 392 |
SSS::::::::SS h::::::h h::::::h aaaaaaa:::::a p:::::p p:::::pe:::::::eeeee::::::e s:::::s ssssss
|
| 393 |
SSSSSS::::S h:::::h h:::::h aa::::::::::::a p:::::p p:::::pe:::::::::::::::::e s::::::s
|
| 394 |
S:::::S h:::::h h:::::h a::::aaaa::::::a p:::::p p:::::pe::::::eeeeeeeeeee s::::::s
|
| 395 |
S:::::S h:::::h h:::::ha::::a a:::::a p:::::p p::::::pe:::::::e ssssss s:::::s
|
| 396 |
SSSSSSS S:::::S h:::::h h:::::ha::::a a:::::a p:::::ppppp:::::::pe::::::::e s:::::ssss::::::s
|
| 397 |
S::::::SSSSSS:::::S h:::::h h:::::ha:::::aaaa::::::a p::::::::::::::::p e::::::::eeeeeeee s::::::::::::::s
|
| 398 |
S:::::::::::::::SS h:::::h h:::::h a::::::::::aa:::ap::::::::::::::pp ee:::::::::::::e s:::::::::::ss
|
| 399 |
SSSSSSSSSSSSSSS hhhhhhh hhhhhhh aaaaaaaaaa aaaap::::::pppppppp eeeeeeeeeeeeee sssssssssss
|
| 400 |
p:::::p
|
| 401 |
p:::::p
|
| 402 |
p:::::::p
|
| 403 |
p:::::::p
|
| 404 |
p:::::::p
|
| 405 |
ppppppppp
|
| 406 |
|
| 407 |
|
| 408 |
*********************************************************************** new : Limits using SHAPES ***********************************/
|
| 409 |
|
| 410 |
|
| 411 |
void limit_shapes_for_systematic_effect(TFile *limfile, string identifier, string mcjzb, string datajzb, int JES,vector<float> binning, TCanvas *limcan) {
|
| 412 |
dout << "Creatig shape templates ... ";
|
| 413 |
if(identifier!="") dout << "for systematic called "<<identifier;
|
| 414 |
dout << endl;
|
| 415 |
int dataormc=mcwithsignal;//this is only for tests - for real life you want dataormc=data !!!
|
| 416 |
if(dataormc!=data) write_warning(__FUNCTION__,"WATCH OUT! Not using data for limits!!!! this is ok for tests, but not ok for anything official!");
|
| 417 |
|
| 418 |
TCut limitnJetcut;
|
| 419 |
if(JES==noJES) limitnJetcut=cutnJets;
|
| 420 |
else {
|
| 421 |
if(JES==JESdown) limitnJetcut=cutnJetsJESdown;
|
| 422 |
if(JES==JESup) limitnJetcut=cutnJetsJESup;
|
| 423 |
}
|
| 424 |
TH1F *ZOSSFP = allsamples.Draw("ZOSSFP",datajzb,binning, "JZB4limits", "events",cutmass&&cutOSSF&&limitnJetcut&&basiccut,dataormc,luminosity);
|
| 425 |
TH1F *ZOSOFP = allsamples.Draw("ZOSOFP",datajzb,binning, "JZB4limits", "events",cutmass&&cutOSOF&&limitnJetcut&&basiccut,dataormc,luminosity);
|
| 426 |
TH1F *ZOSSFN = allsamples.Draw("ZOSSFN","-"+datajzb,binning, "JZB4limits", "events",cutmass&&cutOSSF&&limitnJetcut&&basiccut,dataormc,luminosity);
|
| 427 |
TH1F *ZOSOFN = allsamples.Draw("ZOSOFN","-"+datajzb,binning, "JZB4limits", "events",cutmass&&cutOSOF&&limitnJetcut&&basiccut,dataormc,luminosity);
|
| 428 |
|
| 429 |
TH1F *SBOSSFP;
|
| 430 |
TH1F *SBOSOFP;
|
| 431 |
TH1F *SBOSSFN;
|
| 432 |
TH1F *SBOSOFN;
|
| 433 |
|
| 434 |
TH1F *LZOSSFP = allsamples.Draw("LZOSSFP",mcjzb,binning, "JZB4limits", "events",cutmass&&cutOSSF&&limitnJetcut&&basiccut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 435 |
TH1F *LZOSOFP = allsamples.Draw("LZOSOFP",mcjzb,binning, "JZB4limits", "events",cutmass&&cutOSOF&&limitnJetcut&&basiccut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 436 |
TH1F *LZOSSFN = allsamples.Draw("LZOSSFN","-"+mcjzb,binning, "JZB4limits", "events",cutmass&&cutOSSF&&limitnJetcut&&basiccut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 437 |
TH1F *LZOSOFN = allsamples.Draw("LZOSOFN","-"+mcjzb,binning, "JZB4limits", "events",cutmass&&cutOSOF&&limitnJetcut&&basiccut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 438 |
|
| 439 |
TH1F *LSBOSSFP;
|
| 440 |
TH1F *LSBOSOFP;
|
| 441 |
TH1F *LSBOSSFN;
|
| 442 |
TH1F *LSBOSOFN;
|
| 443 |
|
| 444 |
flag_this_change(__FUNCTION__,__LINE__,false);//PlottingSetup::RestrictToMassPeak
|
| 445 |
if(PlottingSetup::RestrictToMassPeak) {
|
| 446 |
SBOSSFP = allsamples.Draw("SBOSSFP",datajzb,binning, "JZB4limits", "events",cutOSSF&&limitnJetcut&&basiccut&&sidebandcut,dataormc,luminosity);
|
| 447 |
SBOSOFP = allsamples.Draw("SBOSOFP",datajzb,binning, "JZB4limits", "events",cutOSOF&&limitnJetcut&&basiccut&&sidebandcut,dataormc,luminosity);
|
| 448 |
SBOSSFN = allsamples.Draw("SBOSSFN","-"+datajzb,binning, "JZB4limits", "events",cutOSSF&&limitnJetcut&&basiccut&&sidebandcut,dataormc,luminosity);
|
| 449 |
SBOSOFN = allsamples.Draw("SBOSOFN","-"+datajzb,binning, "JZB4limits", "events",cutOSOF&&limitnJetcut&&basiccut&&sidebandcut,dataormc,luminosity);
|
| 450 |
|
| 451 |
LSBOSSFP = allsamples.Draw("LSBOSSFP",mcjzb,binning, "JZB4limits", "events",cutOSSF&&limitnJetcut&&basiccut&&sidebandcut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 452 |
LSBOSOFP = allsamples.Draw("LSBOSOFP",mcjzb,binning, "JZB4limits", "events",cutOSOF&&limitnJetcut&&basiccut&&sidebandcut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 453 |
LSBOSSFN = allsamples.Draw("LSBOSSFN","-"+mcjzb,binning, "JZB4limits", "events",cutOSSF&&limitnJetcut&&basiccut&&sidebandcut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 454 |
LSBOSOFN = allsamples.Draw("LSBOSOFN","-"+mcjzb,binning, "JZB4limits", "events",cutOSOF&&limitnJetcut&&basiccut&&sidebandcut,mc,luminosity,allsamples.FindSample("LM4"));
|
| 455 |
}
|
| 456 |
|
| 457 |
string obsname="data_obs";
|
| 458 |
string predname="background";
|
| 459 |
string signalname="signal";
|
| 460 |
if(identifier!="") {
|
| 461 |
obsname=("data_"+identifier);
|
| 462 |
predname=("background_"+identifier);
|
| 463 |
signalname="signal_"+identifier;
|
| 464 |
}
|
| 465 |
|
| 466 |
TH1F *obs = (TH1F*)ZOSSFP->Clone("observation");
|
| 467 |
obs->SetName(obsname.c_str());
|
| 468 |
obs->Write();
|
| 469 |
TH1F *pred = (TH1F*)ZOSSFN->Clone("prediction");
|
| 470 |
flag_this_change(__FUNCTION__,__LINE__,false);//PlottingSetup::RestrictToMassPeak
|
| 471 |
if(PlottingSetup::RestrictToMassPeak) {
|
| 472 |
pred->Add(ZOSOFP,1.0/3);
|
| 473 |
pred->Add(ZOSOFN,-1.0/3);
|
| 474 |
pred->Add(SBOSSFP,1.0/3);
|
| 475 |
pred->Add(SBOSSFN,-1.0/3);
|
| 476 |
pred->Add(SBOSOFP,1.0/3);
|
| 477 |
pred->Add(SBOSOFN,-1.0/3);
|
| 478 |
} else {
|
| 479 |
pred->Add(ZOSOFP,1.0);
|
| 480 |
pred->Add(ZOSOFN,-1.0);
|
| 481 |
}
|
| 482 |
|
| 483 |
pred->SetName(predname.c_str());
|
| 484 |
pred->Write();
|
| 485 |
|
| 486 |
// TH1F *Lobs = (TH1F*)LZOSSFP->Clone();
|
| 487 |
// TH1F *Lpred = (TH1F*)LZOSSFN->Clone();
|
| 488 |
|
| 489 |
TH1F *Lobs = new TH1F("Lobs","Lobs",binning.size()-1,&binning[0]);
|
| 490 |
TH1F *Lpred = new TH1F("Lpred","Lpred",binning.size()-1,&binning[0]);
|
| 491 |
Lobs->Add(LZOSSFP);
|
| 492 |
Lpred->Add(LZOSSFN);
|
| 493 |
flag_this_change(__FUNCTION__,__LINE__,false);//PlottingSetup::RestrictToMassPeak
|
| 494 |
if(PlottingSetup::RestrictToMassPeak) {
|
| 495 |
Lpred->Add(LZOSOFP,1.0/3);
|
| 496 |
Lpred->Add(LZOSOFN,-1.0/3);
|
| 497 |
Lpred->Add(LSBOSSFP,1.0/3);
|
| 498 |
Lpred->Add(LSBOSSFN,-1.0/3);
|
| 499 |
Lpred->Add(LSBOSOFP,1.0/3);
|
| 500 |
Lpred->Add(LSBOSOFN,-1.0/3);
|
| 501 |
} else {
|
| 502 |
Lpred->Add(LZOSOFP,1.0);
|
| 503 |
Lpred->Add(LZOSOFN,-1.0);
|
| 504 |
}
|
| 505 |
|
| 506 |
TH1F *signal = (TH1F*)Lobs->Clone();
|
| 507 |
signal->Add(Lpred,-1);
|
| 508 |
signal->SetName(signalname.c_str());
|
| 509 |
signal->Write();
|
| 510 |
|
| 511 |
delete Lobs;
|
| 512 |
delete Lpred;
|
| 513 |
|
| 514 |
delete ZOSSFP;
|
| 515 |
delete ZOSOFP;
|
| 516 |
delete ZOSSFN;
|
| 517 |
delete ZOSOFN;
|
| 518 |
|
| 519 |
if(PlottingSetup::RestrictToMassPeak) {
|
| 520 |
delete SBOSSFP;
|
| 521 |
delete SBOSOFP;
|
| 522 |
delete SBOSSFN;
|
| 523 |
delete SBOSOFN;
|
| 524 |
}
|
| 525 |
|
| 526 |
delete LZOSSFP;
|
| 527 |
delete LZOSOFP;
|
| 528 |
delete LZOSSFN;
|
| 529 |
delete LZOSOFN;
|
| 530 |
|
| 531 |
if(PlottingSetup::RestrictToMassPeak) {
|
| 532 |
delete LSBOSSFP;
|
| 533 |
delete LSBOSOFP;
|
| 534 |
delete LSBOSSFN;
|
| 535 |
delete LSBOSOFN;
|
| 536 |
}
|
| 537 |
|
| 538 |
}
|
| 539 |
|
| 540 |
void prepare_datacard(TFile *f) {
|
| 541 |
TH1F *dataob = (TH1F*)f->Get("data_obs");
|
| 542 |
TH1F *signal = (TH1F*)f->Get("signal");
|
| 543 |
TH1F *background = (TH1F*)f->Get("background");
|
| 544 |
|
| 545 |
ofstream datacard;
|
| 546 |
ensure_directory_exists(get_directory()+"/limits");
|
| 547 |
datacard.open ((get_directory()+"/limits/susydatacard.txt").c_str());
|
| 548 |
datacard << "Writing this to a file.\n";
|
| 549 |
datacard << "imax 1\n";
|
| 550 |
datacard << "jmax 1\n";
|
| 551 |
datacard << "kmax *\n";
|
| 552 |
datacard << "---------------\n";
|
| 553 |
datacard << "shapes * * limitfile.root $PROCESS $PROCESS_$SYSTEMATIC\n";
|
| 554 |
datacard << "---------------\n";
|
| 555 |
datacard << "bin 1\n";
|
| 556 |
datacard << "observation "<<dataob->Integral()<<"\n";
|
| 557 |
datacard << "------------------------------\n";
|
| 558 |
datacard << "bin 1 1\n";
|
| 559 |
datacard << "process signal background\n";
|
| 560 |
datacard << "process 0 1\n";
|
| 561 |
datacard << "rate "<<signal->Integral()<<" "<<background->Integral()<<"\n";
|
| 562 |
datacard << "--------------------------------\n";
|
| 563 |
datacard << "lumi lnN 1.10 1.0\n";
|
| 564 |
datacard << "bgnorm lnN 1.00 1.4 uncertainty on our prediction (40%)\n";
|
| 565 |
datacard << "JES shape 1 1 uncertainty on background shape and normalization\n";
|
| 566 |
datacard << "peak shape 1 1 uncertainty on signal resolution. Assume the histogram is a 2 sigma shift, \n";
|
| 567 |
datacard << "# so divide the unit gaussian by 2 before doing the interpolation\n";
|
| 568 |
datacard.close();
|
| 569 |
}
|
| 570 |
|
| 571 |
|
| 572 |
void prepare_limits(string mcjzb, string datajzb, float jzbpeakerrordata, float jzbpeakerrormc, vector<float> jzbbins) {
|
| 573 |
ensure_directory_exists(get_directory()+"/limits");
|
| 574 |
TFile *limfile = new TFile((get_directory()+"/limits/limitfile.root").c_str(),"RECREATE");
|
| 575 |
TCanvas *limcan = new TCanvas("limcan","Canvas for calculating limits");
|
| 576 |
limit_shapes_for_systematic_effect(limfile,"",mcjzb,datajzb,noJES,jzbbins,limcan);
|
| 577 |
limit_shapes_for_systematic_effect(limfile,"peakUp",newjzbexpression(mcjzb,jzbpeakerrormc),newjzbexpression(datajzb,jzbpeakerrordata),noJES,jzbbins,limcan);
|
| 578 |
limit_shapes_for_systematic_effect(limfile,"peakDown",newjzbexpression(mcjzb,-jzbpeakerrormc),newjzbexpression(datajzb,-jzbpeakerrordata),noJES,jzbbins,limcan);
|
| 579 |
limit_shapes_for_systematic_effect(limfile,"JESUp",mcjzb,datajzb,JESup,jzbbins,limcan);
|
| 580 |
limit_shapes_for_systematic_effect(limfile,"JESDown",mcjzb,datajzb,JESdown,jzbbins,limcan);
|
| 581 |
|
| 582 |
prepare_datacard(limfile);
|
| 583 |
limfile->Close();
|
| 584 |
write_info("prepare_limits","limitfile.root and datacard.txt have been generated. You can now use them to calculate limits!");
|
| 585 |
|
| 586 |
}
|
