void Fitter::addGausFitModel(RooArgList *pdfs, RooArgList *coeffs, bool isLast)
{
for( int i=0;i< nGaussians ;++i)
{
RooRealVar *m = new RooRealVar( Form("fitmodel_gaus_mean_%d",i),Form("fitmodel_gaus_mean_%d",i), startMean_[i] );
RooRealVar *s = new RooRealVar( Form("fitmodel_gaus_sigma_%d",i), Form("fitmodel_gaus_sigma_%d",i), startSigma_[i] );
RooGaussian *gaus = new RooGaussian(Form("fitmodel_gaus_g%d",i),Form("fitmodel_gaus_g%d",i),*x_, *m, *s);
vars_[ m->GetName() ] = m;
vars_[ s->GetName() ] = s;
pdfs -> add( *gaus );
if ( (i != nGaussians-1 ) or not isLast)
{
RooRealVar *f = new RooRealVar( Form("fitmodel_gaus_frac_%d",i) ,Form("fitmodel_gaus_frac_%d",i) , startFraction_[i], 0.01,1.0 );
vars_ [ f->GetName() ] = f;
coeffs -> add( *f ) ;
}
}
}
void GenericModel::ImportModelData(Int_t parameter_num, RooArgList* plist)
{
// Import all model datasets corresponding to the defined parameters, ranges & binnings
RooArgList PLIST;
if (plist) PLIST.addClone(*plist);
RooRealVar* par = dynamic_cast<RooRealVar*>(fParameters.at(parameter_num));
RooAbsBinning& bins = par->getBinning();
Int_t N = bins.numBins();
RooRealVar* par_in_list = (RooRealVar*)PLIST.find(par->GetName());
if (!par_in_list) {
PLIST.addClone(RooRealVar(par->GetName(), par->GetTitle(), 0.));
par_in_list = (RooRealVar*)PLIST.find(par->GetName());
}
for (int i = 0; i < N; i++) {
par_in_list->setMax(bins.binHigh(i));
par_in_list->setMin(bins.binLow(i));
par_in_list->setVal(bins.binCenter(i));
if ((parameter_num + 1) < GetNumberOfParameters())
ImportModelData(parameter_num + 1, &PLIST);
else {
AddModelData(PLIST, GetModelDataSet(PLIST));
}
}
}
double TwoBody::printMcmcUpperLimit( double peak, ModelConfig &_mc, std::string filename ){
//
// print out the upper limit on the first Parameter of Interest
//
std::string _legend = "[TwoBody::printMcmcUpperLimit]: ";
char buf[1024];
double _limit = numeric_limits<double>::max();
if (mcInt){
//mc.SetWorkspace(*ws);
RooRealVar * firstPOI = (RooRealVar*) _mc.GetParametersOfInterest()->first();
_limit = mcInt->UpperLimit(*firstPOI);
std::cout << "\n95% upper limit on " << firstPOI->GetName() << " is : " <<
_limit << endl;
if (bMcmcConverged){
sprintf(buf, "%7.1f %7.6f", peak, _limit);
}
else{
sprintf(buf, "# %7.1f %7.6f # MCMC did not converge", peak, _limit);
}
}
else{
sprintf(buf, "# MCMC did not converge");
}
if (filename.size()!=0){
std::ofstream aFile;
// append to file if exists
aFile.open(filename.c_str(), std::ios_base::app);
aFile << buf << std::endl;
// close outfile here so it is safe even if subsequent iterations crash
aFile.close();
}
return _limit;
}
///
/// Set all observables to 'toy' values drawn from the
/// PDF using the current parameter values. A certain number
/// of toys is pregenerated to speed up when doing mulitple toy fits.
///
void PDF_Abs::setObservablesToy()
{
obsValSource = "toy";
if( !pdf ){ cout<< "PDF_Abs::setObservables(): ERROR: pdf not initialized."<<endl; exit(1); }
if ( toyObservables==0 || iToyObs==nToyObs )
{
RooRandom::randomGenerator()->SetSeed(0);
if ( iToyObs==nToyObs ) delete toyObservables;
toyObservables = pdf->generate(*(RooArgSet*)observables, nToyObs);
iToyObs=0;
}
for ( int i=0; i<nObs; i++ )
{
RooRealVar* pObs = (RooRealVar*)((RooArgList*)observables)->at(i);
pObs->setVal(((RooRealVar*)toyObservables->get(iToyObs)->find(pObs->GetName()))->getVal());
}
iToyObs+=1;
}
TTree *dataset2tree(RooDataSet *dataset){
const RooArgSet *args = dataset->get();
if(args==NULL) return NULL;
RooArgList argList(*args);
Double_t variables[50];
Long64_t nEntries= dataset->numEntries();
//nEntries=1;
TTree *tree = new TTree("tree","tree");
tree->SetDirectory(0);
TIterator *it1=NULL;
it1 = argList.createIterator();
int index1=0;
for(RooRealVar *var = (RooRealVar *) it1->Next(); var!=NULL;
var = (RooRealVar *) it1->Next(),index1++){
TString name(var->GetName());
name.ReplaceAll("-","_");
tree->Branch(name, &(variables[index1]), name+"/D");
}
// tree->Print();
for(Long64_t jentry=0; jentry<nEntries; jentry++){
TIterator *it=NULL;
RooArgList argList1(*(dataset->get(jentry)));
it = argList1.createIterator();
//(dataset->get(jentry))->Print();
int index=0;
for(RooRealVar *var = (RooRealVar *) it->Next(); var!=NULL;
var = (RooRealVar *) it->Next(), index++){
variables[index]=var->getVal();
//var->Print();
}
delete it;
tree->Fill();
}
tree->ResetBranchAddresses();
// tree->Print();
return tree;
}
double getChisq(RooAbsData &dat, RooAbsPdf &pdf, RooRealVar &var, bool prt=false) {
// Find total number of events
double nEvt;
double nTot=0.0;
for(int j=0; j<dat.numEntries(); j++) {
dat.get(j);
nEvt=dat.weight();
nTot+=nEvt;
}
// Find chi-squared equivalent 2NLL
//RooRealVar *var=(RooRealVar*)(pdf.getParameters(*dat)->find("CMS_hgg_mass"));
double totNLL=0.0;
double prbSum=0.0;
for(int j=0; j<dat.numEntries(); j++) {
double m=dat.get(j)->getRealValue(var.GetName());
if ( m < var.getMin() || m > var.getMax()) continue;
// Find probability density and hence probability
var.setVal(m);
double prb = var.getBinWidth(0)*pdf.getVal(var);
prbSum+=prb;
dat.get(j);
nEvt=dat.weight();
double mubin=nTot*prb;
double contrib(0.);
if (nEvt < 1) contrib = mubin;
else contrib=mubin-nEvt+nEvt*log(nEvt/mubin);
totNLL+=contrib;
if(prt) cout << "Bin " << j << " prob = " << prb << " nEvt = " << nEvt << ", mu = " << mubin << " contribution " << contrib << endl;
}
totNLL*=2.0;
if(prt) cout << pdf.GetName() << " nTot = " << nTot << " 2NLL constant = " << totNLL << endl;
return totNLL;
}
void OneSidedFrequentistUpperLimitWithBands_intermediate(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
double confidenceLevel=0.95;
// degrade/improve number of pseudo-experiments used to define the confidence belt.
// value of 1 corresponds to default number of toys in the tail, which is 50/(1-confidenceLevel)
double additionalToysFac = 1.;
int nPointsToScan = 30; // number of steps in the parameter of interest
int nToyMC = 100; // number of toys used to define the expected limit and band
TStopwatch t;
t.Start();
/////////////////////////////////////////////////////////////
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
////////////////////////////////////////////////////////////
const char* filename = "";
if (!strcmp(infile,""))
filename = "results/example_combined_GaussExample_model.root";
else
filename = infile;
// Check if example input file exists
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file && strcmp(infile,"")){
cout <<"file not found" << endl;
return;
}
// if default file not found, try to create it
if(!file ){
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
// now try to access the file again
file = TFile::Open(filename);
if(!file){
// if it is still not there, then we can't continue
cout << "Not able to run hist2workspace to create example input" <<endl;
return;
}
/////////////////////////////////////////////////////////////
// Now get the data and workspace
////////////////////////////////////////////////////////////
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
cout << "Found data and ModelConfig:" <<endl;
mc->Print();
/////////////////////////////////////////////////////////////
// Now get the POI for convenience
// you may want to adjust the range of your POI
////////////////////////////////////////////////////////////
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
// firstPOI->setMin(0);
// firstPOI->setMax(10);
/////////////////////////////////////////////
// create and use the FeldmanCousins tool
// to find and plot the 95% confidence interval
// on the parameter of interest as specified
// in the model config
// REMEMBER, we will change the test statistic
// so this is NOT a Feldman-Cousins interval
FeldmanCousins fc(*data,*mc);
fc.SetConfidenceLevel(confidenceLevel);
fc.AdditionalNToysFactor(additionalToysFac); // improve sampling that defines confidence belt
// fc.UseAdaptiveSampling(true); // speed it up a bit, but don't use for expectd limits
fc.SetNBins(nPointsToScan); // set how many points per parameter of interest to scan
fc.CreateConfBelt(true); // save the information in the belt for plotting
/////////////////////////////////////////////
// Feldman-Cousins is a unified limit by definition
// but the tool takes care of a few things for us like which values
// of the nuisance parameters should be used to generate toys.
// so let's just change the test statistic and realize this is
// no longer "Feldman-Cousins" but is a fully frequentist Neyman-Construction.
// ProfileLikelihoodTestStatModified onesided(*mc->GetPdf());
// fc.GetTestStatSampler()->SetTestStatistic(&onesided);
// ((ToyMCSampler*) fc.GetTestStatSampler())->SetGenerateBinned(true);
ToyMCSampler* toymcsampler = (ToyMCSampler*) fc.GetTestStatSampler();
ProfileLikelihoodTestStat* testStat = dynamic_cast<ProfileLikelihoodTestStat*>(toymcsampler->GetTestStatistic());
testStat->SetOneSided(true);
// test speedups:
testStat->SetReuseNLL(true);
// toymcsampler->setUseMultiGen(true); // not fully validated
// Since this tool needs to throw toy MC the PDF needs to be
// extended or the tool needs to know how many entries in a dataset
// per pseudo experiment.
// In the 'number counting form' where the entries in the dataset
// are counts, and not values of discriminating variables, the
// datasets typically only have one entry and the PDF is not
// extended.
if(!mc->GetPdf()->canBeExtended()){
if(data->numEntries()==1)
fc.FluctuateNumDataEntries(false);
else
cout <<"Not sure what to do about this model" <<endl;
}
// We can use PROOF to speed things along in parallel
ProofConfig pc(*w, 4, "",false);
if(mc->GetGlobalObservables()){
cout << "will use global observables for unconditional ensemble"<<endl;
mc->GetGlobalObservables()->Print();
toymcsampler->SetGlobalObservables(*mc->GetGlobalObservables());
}
toymcsampler->SetProofConfig(&pc); // enable proof
// Now get the interval
PointSetInterval* interval = fc.GetInterval();
ConfidenceBelt* belt = fc.GetConfidenceBelt();
// print out the iterval on the first Parameter of Interest
cout << "\n95% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit(*firstPOI) << ", "<<
interval->UpperLimit(*firstPOI) <<"] "<<endl;
// get observed UL and value of test statistic evaluated there
RooArgSet tmpPOI(*firstPOI);
double observedUL = interval->UpperLimit(*firstPOI);
firstPOI->setVal(observedUL);
double obsTSatObsUL = fc.GetTestStatSampler()->EvaluateTestStatistic(*data,tmpPOI);
// Ask the calculator which points were scanned
RooDataSet* parameterScan = (RooDataSet*) fc.GetPointsToScan();
RooArgSet* tmpPoint;
// make a histogram of parameter vs. threshold
TH1F* histOfThresholds = new TH1F("histOfThresholds","",
parameterScan->numEntries(),
firstPOI->getMin(),
firstPOI->getMax());
histOfThresholds->GetXaxis()->SetTitle(firstPOI->GetName());
histOfThresholds->GetYaxis()->SetTitle("Threshold");
// loop through the points that were tested and ask confidence belt
// what the upper/lower thresholds were.
// For FeldmanCousins, the lower cut off is always 0
for(Int_t i=0; i<parameterScan->numEntries(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
double arMax = belt->GetAcceptanceRegionMax(*tmpPoint);
double poiVal = tmpPoint->getRealValue(firstPOI->GetName()) ;
histOfThresholds->Fill(poiVal,arMax);
}
TCanvas* c1 = new TCanvas();
c1->Divide(2);
c1->cd(1);
histOfThresholds->SetMinimum(0);
histOfThresholds->Draw();
c1->cd(2);
/////////////////////////////////////////////////////////////
// Now we generate the expected bands and power-constriant
////////////////////////////////////////////////////////////
// First: find parameter point for mu=0, with conditional MLEs for nuisance parameters
RooAbsReal* nll = mc->GetPdf()->createNLL(*data);
RooAbsReal* profile = nll->createProfile(*mc->GetParametersOfInterest());
firstPOI->setVal(0.);
profile->getVal(); // this will do fit and set nuisance parameters to profiled values
RooArgSet* poiAndNuisance = new RooArgSet();
if(mc->GetNuisanceParameters())
poiAndNuisance->add(*mc->GetNuisanceParameters());
poiAndNuisance->add(*mc->GetParametersOfInterest());
w->saveSnapshot("paramsToGenerateData",*poiAndNuisance);
RooArgSet* paramsToGenerateData = (RooArgSet*) poiAndNuisance->snapshot();
cout << "\nWill use these parameter points to generate pseudo data for bkg only" << endl;
paramsToGenerateData->Print("v");
double CLb=0;
double CLbinclusive=0;
// Now we generate background only and find distribution of upper limits
TH1F* histOfUL = new TH1F("histOfUL","",100,0,firstPOI->getMax());
histOfUL->GetXaxis()->SetTitle("Upper Limit (background only)");
histOfUL->GetYaxis()->SetTitle("Entries");
for(int imc=0; imc<nToyMC; ++imc){
// set parameters back to values for generating pseudo data
w->loadSnapshot("paramsToGenerateData");
// in 5.30 there is a nicer way to generate toy data & randomize global obs
RooAbsData* toyData = toymcsampler->GenerateToyData(*paramsToGenerateData);
// get test stat at observed UL in observed data
firstPOI->setVal(observedUL);
double toyTSatObsUL = fc.GetTestStatSampler()->EvaluateTestStatistic(*toyData,tmpPOI);
// toyData->get()->Print("v");
// cout <<"obsTSatObsUL " <<obsTSatObsUL << "toyTS " << toyTSatObsUL << endl;
if(obsTSatObsUL < toyTSatObsUL) // (should be checked)
CLb+= (1.)/nToyMC;
if(obsTSatObsUL <= toyTSatObsUL) // (should be checked)
CLbinclusive+= (1.)/nToyMC;
// loop over points in belt to find upper limit for this toy data
double thisUL = 0;
for(Int_t i=0; i<parameterScan->numEntries(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
double arMax = belt->GetAcceptanceRegionMax(*tmpPoint);
firstPOI->setVal( tmpPoint->getRealValue(firstPOI->GetName()) );
double thisTS = fc.GetTestStatSampler()->EvaluateTestStatistic(*toyData,tmpPOI);
if(thisTS<=arMax){
thisUL = firstPOI->getVal();
} else{
break;
}
}
histOfUL->Fill(thisUL);
delete toyData;
}
histOfUL->Draw();
c1->SaveAs("one-sided_upper_limit_output.pdf");
// if you want to see a plot of the sampling distribution for a particular scan point:
// Now find bands and power constraint
Double_t* bins = histOfUL->GetIntegral();
TH1F* cumulative = (TH1F*) histOfUL->Clone("cumulative");
cumulative->SetContent(bins);
double band2sigDown=0, band1sigDown=0, bandMedian=0, band1sigUp=0,band2sigUp=0;
for(int i=1; i<=cumulative->GetNbinsX(); ++i){
if(bins[i]<RooStats::SignificanceToPValue(2))
band2sigDown=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(1))
band1sigDown=cumulative->GetBinCenter(i);
if(bins[i]<0.5)
bandMedian=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(-1))
band1sigUp=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(-2))
band2sigUp=cumulative->GetBinCenter(i);
}
t.Stop();
t.Print();
cout << "-2 sigma band " << band2sigDown << endl;
cout << "-1 sigma band " << band1sigDown << endl;
cout << "median of band " << bandMedian << " [Power Constriant)]" << endl;
cout << "+1 sigma band " << band1sigUp << endl;
cout << "+2 sigma band " << band2sigUp << endl;
// print out the iterval on the first Parameter of Interest
cout << "\nobserved 95% upper-limit "<< interval->UpperLimit(*firstPOI) <<endl;
cout << "CLb strict [P(toy>obs|0)] for observed 95% upper-limit "<< CLb <<endl;
cout << "CLb inclusive [P(toy>=obs|0)] for observed 95% upper-limit "<< CLbinclusive <<endl;
delete profile;
delete nll;
}
void THSEventsPDF::adjustBinning(Int_t* offset1) const
{
RooRealVar* xvar = fx_off ;
if (!dynamic_cast<RooRealVar*>(xvar)) {
coutE(InputArguments) << "RooDataHist::adjustBinning(" << GetName() << ") ERROR: dimension " << xvar->GetName() << " must be real" << endl ;
assert(0) ;
}
Double_t xlo = xvar->getMin() ;
Double_t xhi = xvar->getMax() ;
//adjust bin range limits with new scale parameter
//cout<<scale<<" "<<fMean<<" "<<xlo<<" "<<xhi<<endl;
xlo=(xlo-fMean)/scale+fMean;
xhi=(xhi-fMean)/scale+fMean;
if(xvar->getBinning().lowBound()==xlo&&xvar->getBinning().highBound()==xhi) return;
xvar->setRange(xlo,xhi) ;
// Int_t xmin(0) ;
// cout<<"THSEventsPDF::adjustBinning( "<<xlo <<" "<<xhi<<endl;
//now adjust fitting range to bin limits??Possibly not
if (fRHist->GetXaxis()->GetXbins()->GetArray()) {
RooBinning xbins(fRHist->GetNbinsX(),fRHist->GetXaxis()->GetXbins()->GetArray()) ;
Double_t tolerance = 1e-6*xbins.averageBinWidth() ;
// Adjust xlo/xhi to nearest boundary
Double_t xloAdj = xbins.binLow(xbins.binNumber(xlo+tolerance)) ;
Double_t xhiAdj = xbins.binHigh(xbins.binNumber(xhi-tolerance)) ;
xbins.setRange(xloAdj,xhiAdj) ;
xvar->setBinning(xbins) ;
if (fabs(xloAdj-xlo)>tolerance||fabs(xhiAdj-xhi)<tolerance) {
coutI(DataHandling) << "RooDataHist::adjustBinning(" << GetName() << "): fit range of variable " << xvar->GetName() << " expanded to nearest bin boundaries: ["
<< xlo << "," << xhi << "] --> [" << xloAdj << "," << xhiAdj << "]" << endl ;
}
} else {
RooBinning xbins(fRHist->GetXaxis()->GetXmin(),fRHist->GetXaxis()->GetXmax()) ;
xbins.addUniform(fRHist->GetNbinsX(),fRHist->GetXaxis()->GetXmin(),fRHist->GetXaxis()->GetXmax()) ;
Double_t tolerance = 1e-6*xbins.averageBinWidth() ;
// Adjust xlo/xhi to nearest boundary
Double_t xloAdj = xbins.binLow(xbins.binNumber(xlo+tolerance)) ;
Double_t xhiAdj = xbins.binHigh(xbins.binNumber(xhi-tolerance)) ;
xbins.setRange(xloAdj,xhiAdj) ;
xvar->setRange(xloAdj,xhiAdj) ;
//xvar->setRange(xlo,xhi) ;
}
return;
}
///
/// Find the global minimum in a more thorough way.
/// First fit with external start parameters, then
/// for each parameter that starts with "d" or "r" (typically angles and ratios):
/// - at upper scan range, rest at start parameters
/// - at lower scan range, rest at start parameters
/// This amounts to a maximum of 1+2^n fits, where n is the number
/// of parameters to be varied.
///
/// \param w Workspace holding the pdf.
/// \param name Name of the pdf without leading "pdf_".
/// \param forceVariables Apply the force method for these variables only. Format
/// "var1,var2,var3," (list must end with comma). Default is to apply for all angles,
/// all ratios except rD_k3pi and rD_kpi, and the k3pi coherence factor.
///
RooFitResult* Utils::fitToMinForce(RooWorkspace *w, TString name, TString forceVariables)
{
bool debug = true;
TString parsName = "par_"+name;
TString obsName = "obs_"+name;
TString pdfName = "pdf_"+name;
RooFitResult *r = 0;
int printlevel = -1;
RooMsgService::instance().setGlobalKillBelow(ERROR);
// save start parameters
if ( !w->set(parsName) ){
cout << "MethodProbScan::scan2d() : ERROR : parsName not found: " << parsName << endl;
exit(1);
}
RooDataSet *startPars = new RooDataSet("startParsForce", "startParsForce", *w->set(parsName));
startPars->add(*w->set(parsName));
// set up parameters and ranges
RooArgList *varyPars = new RooArgList();
TIterator* it = w->set(parsName)->createIterator();
while ( RooRealVar* p = (RooRealVar*)it->Next() )
{
if ( p->isConstant() ) continue;
if ( forceVariables=="" && ( false
|| TString(p->GetName()).BeginsWith("d") ///< use these variables
// || TString(p->GetName()).BeginsWith("r")
|| TString(p->GetName()).BeginsWith("k")
|| TString(p->GetName()) == "g"
) && ! (
TString(p->GetName()) == "rD_k3pi" ///< don't use these
|| TString(p->GetName()) == "rD_kpi"
// || TString(p->GetName()) == "dD_kpi"
|| TString(p->GetName()) == "d_dk"
|| TString(p->GetName()) == "d_dsk"
))
{
varyPars->add(*p);
}
else if ( forceVariables.Contains(TString(p->GetName())+",") )
{
varyPars->add(*p);
}
}
delete it;
int nPars = varyPars->getSize();
if ( debug ) cout << "Utils::fitToMinForce() : nPars = " << nPars << " => " << pow(2.,nPars) << " fits" << endl;
if ( debug ) cout << "Utils::fitToMinForce() : varying ";
if ( debug ) varyPars->Print();
//////////
r = fitToMinBringBackAngles(w->pdf(pdfName), false, printlevel);
//////////
int nErrors = 0;
// We define a binary mask where each bit corresponds
// to parameter at max or at min.
for ( int i=0; i<pow(2.,nPars); i++ )
{
if ( debug ) cout << "Utils::fitToMinForce() : fit " << i << " \r" << flush;
setParameters(w, parsName, startPars->get(0));
for ( int ip=0; ip<nPars; ip++ )
{
RooRealVar *p = (RooRealVar*)varyPars->at(ip);
float oldMin = p->getMin();
float oldMax = p->getMax();
setLimit(w, p->GetName(), "force");
if ( i/(int)pow(2.,ip) % 2==0 ) { p->setVal(p->getMin()); }
if ( i/(int)pow(2.,ip) % 2==1 ) { p->setVal(p->getMax()); }
p->setRange(oldMin, oldMax);
}
// check if start parameters are sensible, skip if they're not
double startParChi2 = getChi2(w->pdf(pdfName));
if ( startParChi2>2000 ){
nErrors += 1;
continue;
}
// refit
RooFitResult *r2 = fitToMinBringBackAngles(w->pdf(pdfName), false, printlevel);
// In case the initial fit failed, accept the second one.
// If both failed, still select the second one and hope the
// next fit succeeds.
if ( !(r->edm()<1 && r->covQual()==3) ){
delete r;
r = r2;
}
else if ( r2->edm()<1 && r2->covQual()==3 && r2->minNll()<r->minNll() ){
// better minimum found!
delete r;
r = r2;
}
else{
delete r2;
}
}
if ( debug ) cout << endl;
if ( debug ) cout << "Utils::fitToMinForce() : nErrors = " << nErrors << endl;
RooMsgService::instance().setGlobalKillBelow(INFO);
// (re)set to best parameters
setParameters(w, parsName, r);
delete startPars;
return r;
}
void combinedWorkspace_4WS(const char* name_pbpb_pass="******", const char* name_pbpb_fail="fitresult_pbpb_fail.root", const char* name_pp_pass="******", const char* name_pp_fail="fitresult_pp_fail.root", const char* name_out="fitresult_combo.root", const float systval = 0., const char* subDirName ="wsTest", int nCPU=2){
// subdir: Directory to save workspaces under currentPATH/CombinedWorkspaces/subDir/
// set things silent
gErrorIgnoreLevel=kError;
RooMsgService::instance().setGlobalKillBelow(RooFit::ERROR);
bool dosyst = (systval > 0.);
TString nameOut(name_out);
RooWorkspace * ws = test_combine_4WS(name_pbpb_pass, name_pp_pass, name_pbpb_fail, name_pp_fail, false, nCPU);
RooAbsData * data = ws->data("dOS_DATA");
RooRealVar* RFrac2Svs1S_PbPbvsPP_P = ws->var("RFrac2Svs1S_PbPbvsPP_P");
RooRealVar* leftEdge = new RooRealVar("leftEdge","leftEdge",-10);
RooRealVar* rightEdge = new RooRealVar("rightEdge","rightEdge",10);
RooGenericPdf step("step", "step", "(@0 >= @1) && (@0 < @2)", RooArgList(*RFrac2Svs1S_PbPbvsPP_P, *leftEdge, *rightEdge));
ws->import(step);
ws->factory( "Uniform::flat(RFrac2Svs1S_PbPbvsPP_P)" );
// systematics
if (dosyst) {
ws->factory( Form("kappa_syst[%f]",systval) );
ws->factory( "expr::alpha_syst('kappa_syst*beta_syst',kappa_syst,beta_syst[0,-5,5])" );
ws->factory( "Gaussian::constr_syst(beta_syst,glob_syst[0,-5,5],1)" );
// add systematics into the double ratio
ws->factory( "expr::RFrac2Svs1S_PbPbvsPP_P_syst('@0+@1',RFrac2Svs1S_PbPbvsPP_P,alpha_syst)" );
// build the pbpb pdf
RooRealVar* effjpsi_pp_P = (RooRealVar*)ws->var("effjpsi_pp_P");
RooRealVar* effpsip_pp_P = (RooRealVar*)ws->var("effpsip_pp_P");
RooRealVar* effjpsi_pp_NP = (RooRealVar*)ws->var("effjpsi_pp_NP");
Double_t Npsi2SPbPbPass = npsip_pbpb_pass_from_doubleratio_prompt(ws, RooArgList(*effjpsi_pp_P,*effpsip_pp_P,*effjpsi_pp_NP),true); // Create and import N_Psi2S_PbPb_pass_syst
ws->factory( "SUM::pdfMASS_Tot_PbPb_pass_syst(N_Jpsi_PbPb_pass * pdfMASS_Jpsi_PbPb_pass, N_Psi2S_PbPb_pass_syst * pdfMASS_Psi2S_PbPb_pass, N_Bkg_PbPb_pass * pdfMASS_Bkg_PbPb_pass)" );
ws->factory( "PROD::pdfMASS_Tot_PbPb_pass_constr(pdfMASS_Tot_PbPb_pass_syst,constr_syst)" );
// build the combined pdf
ws->factory("SIMUL::simPdf_syst_noconstr(sample,PbPb_pass=pdfMASS_Tot_PbPb_pass_syst,PbPb_fail=pdfMASS_Tot_PbPb_fail,PP_pass=pdfMASS_Tot_PP_pass,PP_fail=pdfMASS_Tot_PP_fail)");
RooSimultaneous *simPdf = (RooSimultaneous*) ws->pdf("simPdf_syst_noconstr");
RooGaussian *constr_syst = (RooGaussian*) ws->pdf("constr_syst");
RooProdPdf *simPdf_constr = new RooProdPdf("simPdf_syst","simPdf_syst",RooArgSet(*simPdf,*constr_syst));
ws->import(*simPdf_constr);
} else {
ws->factory("SIMUL::simPdf_syst(sample,PbPb_pass=pdfMASS_Tot_PbPb_pass,PbPb_fail=pdfMASS_Tot_PbPb_fail,PP_pass=pdfMASS_Tot_PP_pass,PP_fail=pdfMASS_Tot_PP_fail)");
}
ws->Print();
if (dosyst) ws->var("beta_syst")->setConstant(kFALSE);
/////////////////////////////////////////////////////////////////////
RooRealVar * pObs = ws->var("invMass"); // get the pointer to the observable
RooArgSet obs("observables");
obs.add(*pObs);
obs.add( *ws->cat("sample"));
// /////////////////////////////////////////////////////////////////////
if (dosyst) ws->var("glob_syst")->setConstant(true);
RooArgSet globalObs("global_obs");
if (dosyst) globalObs.add( *ws->var("glob_syst") );
// ws->Print();
RooArgSet poi("poi");
poi.add( *ws->var("RFrac2Svs1S_PbPbvsPP_P") );
// create set of nuisance parameters
RooArgSet nuis("nuis");
if (dosyst) nuis.add( *ws->var("beta_syst") );
// set parameters constant
RooArgSet allVars = ws->allVars();
TIterator* it = allVars.createIterator();
RooRealVar *theVar = (RooRealVar*) it->Next();
while (theVar) {
TString varname(theVar->GetName());
// if (varname != "RFrac2Svs1S_PbPbvsPP"
// && varname != "invMass"
// && varname != "sample"
// )
// theVar->setConstant();
if ( varname.Contains("f_Jpsi_PP") || varname.Contains("f_Jpsi_PbPb") ||
varname.Contains("rSigma21_Jpsi_PP") ||
varname.Contains("m_Jpsi_PP") || varname.Contains("m_Jpsi_PbPb") ||
varname.Contains("sigma1_Jpsi_PP") || varname.Contains("sigma1_Jpsi_PbPb") ||
(varname.Contains("lambda")) ||
(varname.Contains("_fail") && !varname.Contains("RFrac2Svs1S")))
{
theVar->setConstant();
}
if (varname=="glob_syst"
|| varname=="beta_syst"
) {
cout << varname << endl;
theVar->setConstant(!dosyst);
}
theVar = (RooRealVar*) it->Next();
}
// create signal+background Model Config
RooStats::ModelConfig sbHypo("SbHypo");
sbHypo.SetWorkspace( *ws );
sbHypo.SetPdf( *ws->pdf("simPdf_syst") );
sbHypo.SetObservables( obs );
sbHypo.SetGlobalObservables( globalObs );
sbHypo.SetParametersOfInterest( poi );
sbHypo.SetNuisanceParameters( nuis );
sbHypo.SetPriorPdf( *ws->pdf("step") ); // this is optional
/////////////////////////////////////////////////////////////////////
RooAbsReal * pNll = sbHypo.GetPdf()->createNLL( *data,NumCPU(nCPU) );
RooMinuit(*pNll).migrad(); // minimize likelihood wrt all parameters before making plots
if (controlPlots)
{
RooPlot *framepoi = ((RooRealVar *)poi.first())->frame(Bins(10),Range(0.,1),Title("LL and profileLL in RFrac2Svs1S_PbPbvsPP_P"));
pNll->plotOn(framepoi,ShiftToZero());
framepoi->SetMinimum(0);
framepoi->SetMaximum(10);
TCanvas *cpoi = new TCanvas();
cpoi->cd(); framepoi->Draw();
cpoi->SaveAs("cpoi.pdf");
}
((RooRealVar *)poi.first())->setMin(0.);
RooArgSet * pPoiAndNuisance = new RooArgSet("poiAndNuisance");
pPoiAndNuisance->add( nuis );
pPoiAndNuisance->add( poi );
sbHypo.SetSnapshot(*pPoiAndNuisance);
if (controlPlots)
{
RooPlot* xframeSB_PP_pass = pObs->frame(Title("SBhypo_PP_pass"));
data->plotOn(xframeSB_PP_pass,Cut("sample==sample::PP_pass"));
RooAbsPdf *pdfSB_PP_pass = sbHypo.GetPdf();
RooCategory *sample = ws->cat("sample");
pdfSB_PP_pass->plotOn(xframeSB_PP_pass,Slice(*sample,"PP_pass"),ProjWData(*sample,*data));
TCanvas *c1 = new TCanvas();
c1->cd(); xframeSB_PP_pass->Draw();
c1->SaveAs("c1.pdf");
RooPlot* xframeSB_PP_fail = pObs->frame(Title("SBhypo_PP_fail"));
data->plotOn(xframeSB_PP_fail,Cut("sample==sample::PP_fail"));
RooAbsPdf *pdfSB_PP_fail = sbHypo.GetPdf();
pdfSB_PP_fail->plotOn(xframeSB_PP_fail,Slice(*sample,"PP_fail"),ProjWData(*sample,*data));
TCanvas *c2 = new TCanvas();
c2->cd(); xframeSB_PP_fail->Draw();
c2->SaveAs("c1.pdf");
RooPlot* xframeB_PbPb_pass = pObs->frame(Title("SBhypo_PbPb_pass"));
data->plotOn(xframeB_PbPb_pass,Cut("sample==sample::PbPb_pass"));
RooAbsPdf *pdfB_PbPb_pass = sbHypo.GetPdf();
pdfB_PbPb_pass->plotOn(xframeB_PbPb_pass,Slice(*sample,"PbPb_pass"),ProjWData(*sample,*data));
TCanvas *c3 = new TCanvas();
c3->cd(); xframeB_PbPb_pass->Draw();
c3->SetLogy();
c3->SaveAs("c2.pdf");
RooPlot* xframeB_PbPb_fail = pObs->frame(Title("SBhypo_PbPb_fail"));
data->plotOn(xframeB_PbPb_fail,Cut("sample==sample::PbPb_fail"));
RooAbsPdf *pdfB_PbPb_fail = sbHypo.GetPdf();
pdfB_PbPb_fail->plotOn(xframeB_PbPb_fail,Slice(*sample,"PbPb_fail"),ProjWData(*sample,*data));
TCanvas *c4 = new TCanvas();
c4->cd(); xframeB_PbPb_fail->Draw();
c4->SetLogy();
c4->SaveAs("c2.pdf");
}
delete pNll;
delete pPoiAndNuisance;
ws->import( sbHypo );
/////////////////////////////////////////////////////////////////////
RooStats::ModelConfig bHypo = sbHypo;
bHypo.SetName("BHypo");
bHypo.SetWorkspace(*ws);
pNll = bHypo.GetPdf()->createNLL( *data,NumCPU(nCPU) );
// RooMinuit(*pNll).migrad(); // minimize likelihood wrt all parameters before making plots
RooArgSet poiAndGlobalObs("poiAndGlobalObs");
poiAndGlobalObs.add( poi );
poiAndGlobalObs.add( globalObs );
RooAbsReal * pProfile = pNll->createProfile( poiAndGlobalObs ); // do not profile POI and global observables
((RooRealVar *)poi.first())->setVal( 0 ); // set RFrac2Svs1S_PbPbvsPP=0 here
pProfile->getVal(); // this will do fit and set nuisance parameters to profiled values
pPoiAndNuisance = new RooArgSet( "poiAndNuisance" );
pPoiAndNuisance->add( nuis );
pPoiAndNuisance->add( poi );
bHypo.SetSnapshot(*pPoiAndNuisance);
delete pNll;
delete pPoiAndNuisance;
// import model config into workspace
bHypo.SetWorkspace(*ws);
ws->import( bHypo );
/////////////////////////////////////////////////////////////////////
ws->Print();
bHypo.Print();
sbHypo.Print();
// save workspace to file
string mainDIR = gSystem->ExpandPathName(gSystem->pwd());
string wsDIR = mainDIR + "/CombinedWorkspaces/";
string ssubDirName="";
if (subDirName) ssubDirName.append(subDirName);
string subDIR = wsDIR + ssubDirName;
void * dirp = gSystem->OpenDirectory(wsDIR.c_str());
if (dirp) gSystem->FreeDirectory(dirp);
else gSystem->mkdir(wsDIR.c_str(), kTRUE);
void * dirq = gSystem->OpenDirectory(subDIR.c_str());
if (dirq) gSystem->FreeDirectory(dirq);
else gSystem->mkdir(subDIR.c_str(), kTRUE);
const char* saveName = Form("%s/%s",subDIR.c_str(),nameOut.Data());
ws->writeToFile(saveName);
}
void StandardProfileLikelihoodDemo(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
double confLevel = optPL.confLevel;
double nScanPoints = optPL.nScanPoints;
bool plotAsTF1 = optPL.plotAsTF1;
double poiXMin = optPL.poiMinPlot;
double poiXMax = optPL.poiMaxPlot;
bool doHypoTest = optPL.doHypoTest;
double nullParamValue = optPL.nullValue;
// -------------------------------------------------------
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
// -------------------------------------------------------
// Tutorial starts here
// -------------------------------------------------------
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
// ---------------------------------------------
// create and use the ProfileLikelihoodCalculator
// to find and plot the 95% confidence interval
// on the parameter of interest as specified
// in the model config
ProfileLikelihoodCalculator pl(*data,*mc);
pl.SetConfidenceLevel(confLevel); // 95% interval
LikelihoodInterval* interval = pl.GetInterval();
// print out the interval on the first Parameter of Interest
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
cout << "\n>>>> RESULT : " << confLevel*100 << "% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit(*firstPOI) << ", "<<
interval->UpperLimit(*firstPOI) <<"]\n "<<endl;
// make a plot
cout << "making a plot of the profile likelihood function ....(if it is taking a lot of time use less points or the TF1 drawing option)\n";
LikelihoodIntervalPlot plot(interval);
plot.SetNPoints(nScanPoints); // do not use too many points, it could become very slow for some models
if (poiXMin < poiXMax) plot.SetRange(poiXMin, poiXMax);
TString opt;
if (plotAsTF1) opt += TString("tf1");
plot.Draw(opt); // use option TF1 if too slow (plot.Draw("tf1")
// if requested perform also an hypothesis test for the significance
if (doHypoTest) {
RooArgSet nullparams("nullparams");
nullparams.addClone(*firstPOI);
nullparams.setRealValue(firstPOI->GetName(), nullParamValue);
pl.SetNullParameters(nullparams);
std::cout << "Perform Test of Hypothesis : null Hypothesis is " << firstPOI->GetName() << nullParamValue << std::endl;
auto result = pl.GetHypoTest();
std::cout << "\n>>>> Hypotheis Test Result \n";
result->Print();
}
}
void rf403_weightedevts()
{
// C r e a t e o b s e r v a b l e a n d u n w e i g h t e d d a t a s e t
// -------------------------------------------------------------------------------
// Declare observable
RooRealVar x("x","x",-10,10) ;
x.setBins(40) ;
// Construction a uniform pdf
RooPolynomial p0("px","px",x) ;
// Sample 1000 events from pdf
RooDataSet* data = p0.generate(x,1000) ;
// C a l c u l a t e w e i g h t a n d m a k e d a t a s e t w e i g h t e d
// -----------------------------------------------------------------------------------
// Construct formula to calculate (fake) weight for events
RooFormulaVar wFunc("w","event weight","(x*x+10)",x) ;
// Add column with variable w to previously generated dataset
RooRealVar* w = (RooRealVar*) data->addColumn(wFunc) ;
// Dataset d is now a dataset with two observable (x,w) with 1000 entries
data->Print() ;
// Instruct dataset wdata in interpret w as event weight rather than as observable
RooDataSet wdata(data->GetName(),data->GetTitle(),data,*data->get(),0,w->GetName()) ;
// Dataset d is now a dataset with one observable (x) with 1000 entries and a sum of weights of ~430K
wdata.Print() ;
// U n b i n n e d M L f i t t o w e i g h t e d d a t a
// ---------------------------------------------------------------
// Construction quadratic polynomial pdf for fitting
RooRealVar a0("a0","a0",1) ;
RooRealVar a1("a1","a1",0,-1,1) ;
RooRealVar a2("a2","a2",1,0,10) ;
RooPolynomial p2("p2","p2",x,RooArgList(a0,a1,a2),0) ;
// Fit quadratic polynomial to weighted data
// NOTE: A plain Maximum likelihood fit to weighted data does in general
// NOT result in correct error estimates, unless individual
// event weights represent Poisson statistics themselves.
//
// Fit with 'wrong' errors
RooFitResult* r_ml_wgt = p2.fitTo(wdata,Save()) ;
// A first order correction to estimated parameter errors in an
// (unbinned) ML fit can be obtained by calculating the
// covariance matrix as
//
// V' = V C-1 V
//
// where V is the covariance matrix calculated from a fit
// to -logL = - sum [ w_i log f(x_i) ] and C is the covariance
// matrix calculated from -logL' = -sum [ w_i^2 log f(x_i) ]
// (i.e. the weights are applied squared)
//
// A fit in this mode can be performed as follows:
RooFitResult* r_ml_wgt_corr = p2.fitTo(wdata,Save(),SumW2Error(kTRUE)) ;
// P l o t w e i g h e d d a t a a n d f i t r e s u l t
// ---------------------------------------------------------------
// Construct plot frame
RooPlot* frame = x.frame(Title("Unbinned ML fit, binned chi^2 fit to weighted data")) ;
// Plot data using sum-of-weights-squared error rather than Poisson errors
wdata.plotOn(frame,DataError(RooAbsData::SumW2)) ;
// Overlay result of 2nd order polynomial fit to weighted data
p2.plotOn(frame) ;
// M L F i t o f p d f t o e q u i v a l e n t u n w e i g h t e d d a t a s e t
// -----------------------------------------------------------------------------------------
// Construct a pdf with the same shape as p0 after weighting
RooGenericPdf genPdf("genPdf","x*x+10",x) ;
// Sample a dataset with the same number of events as data
RooDataSet* data2 = genPdf.generate(x,1000) ;
// Sample a dataset with the same number of weights as data
RooDataSet* data3 = genPdf.generate(x,43000) ;
// Fit the 2nd order polynomial to both unweighted datasets and save the results for comparison
RooFitResult* r_ml_unw10 = p2.fitTo(*data2,Save()) ;
RooFitResult* r_ml_unw43 = p2.fitTo(*data3,Save()) ;
// C h i 2 f i t o f p d f t o b i n n e d w e i g h t e d d a t a s e t
// ------------------------------------------------------------------------------------
// Construct binned clone of unbinned weighted dataset
RooDataHist* binnedData = wdata.binnedClone() ;
binnedData->Print("v") ;
// Perform chi2 fit to binned weighted dataset using sum-of-weights errors
//
// NB: Within the usual approximations of a chi2 fit, a chi2 fit to weighted
// data using sum-of-weights-squared errors does give correct error
// estimates
RooChi2Var chi2("chi2","chi2",p2,*binnedData,DataError(RooAbsData::SumW2)) ;
RooMinuit m(chi2) ;
m.migrad() ;
m.hesse() ;
// Plot chi^2 fit result on frame as well
RooFitResult* r_chi2_wgt = m.save() ;
p2.plotOn(frame,LineStyle(kDashed),LineColor(kRed)) ;
// C o m p a r e f i t r e s u l t s o f c h i 2 , M L f i t s t o ( u n ) w e i g h t e d d a t a
// ---------------------------------------------------------------------------------------------------------------
// Note that ML fit on 1Kevt of weighted data is closer to result of ML fit on 43Kevt of unweighted data
// than to 1Kevt of unweighted data, whereas the reference chi^2 fit with SumW2 error gives a result closer to
// that of an unbinned ML fit to 1Kevt of unweighted data.
cout << "==> ML Fit results on 1K unweighted events" << endl ;
r_ml_unw10->Print() ;
cout << "==> ML Fit results on 43K unweighted events" << endl ;
r_ml_unw43->Print() ;
cout << "==> ML Fit results on 1K weighted events with a summed weight of 43K" << endl ;
r_ml_wgt->Print() ;
cout << "==> Corrected ML Fit results on 1K weighted events with a summed weight of 43K" << endl ;
r_ml_wgt_corr->Print() ;
cout << "==> Chi2 Fit results on 1K weighted events with a summed weight of 43K" << endl ;
r_chi2_wgt->Print() ;
new TCanvas("rf403_weightedevts","rf403_weightedevts",600,600) ;
gPad->SetLeftMargin(0.15) ; frame->GetYaxis()->SetTitleOffset(1.8) ; frame->Draw() ;
}
void fit_mass(TString fileN="") {//suffix added before file extension, e.g., '.pdf'
TString placeholder;//to add strings before using them, e.g., for saving text files
gROOT->SetBatch(kTRUE);
gROOT->ProcessLine(".x /afs/cern.ch/user/m/mwilkins/cmtuser/src/lhcbStyle.C");
// gStyle->SetPadTickX(1);
// gStyle->SetPadTickY(1);
// gStyle->SetPadLeftMargin(0.15);
// gStyle->SetTextSize(0.3);
// //open file and get histogram
// TFile *inHistos = new TFile("/afs/cern.ch/work/m/mwilkins/Lb2JpsiLtr/data/histos_data.root", "READ");
// TH1F * h100 = (TH1F*)inHistos->Get("h70");
// cout<<"data histogram gotten"<<endl;
//unbinned
TFile *hastree = new TFile("/afs/cern.ch/work/m/mwilkins/Lb2JpsiLtr/data/cutfile_Optimized.root", "READ");
TTree * h100 = (TTree*)hastree->Get("mytree");
cout<<"tree gotten"<<endl;
TFile *SMChistos= new TFile("/afs/cern.ch/work/m/mwilkins/Lb2JpsiLtr/MC/withKScut/histos_SMCfile_fullMC.root", "READ");
cout<<"SMC file opened"<<endl;
TH1F *SMCh = (TH1F*)SMChistos->Get("h00");
cout<<"SMC hist gotten"<<endl;
RooRealVar *mass = new RooRealVar("Bs_LOKI_MASS_JpsiConstr","m(J/#psi #Lambda)",4100,6100,"MeV");
mass->setRange("bkg1",4300,4800);
mass->setRange("bkg2",5700,5950);
mass->setRange("bkg3",4300,5500);
mass->setRange("bkg4",5100,5500);
mass->setRange("L",5350,5950);
mass->setRange("tot",4300,5950);
cout<<"mass declared"<<endl;
// RooDataHist *data = new RooDataHist("data","1D",RooArgList(*mass),h100);
//unbinned
RooDataSet *data = new RooDataSet("data","1D",h100,*mass);
cout<<"data declared"<<endl;
RooDataHist *SMC = new RooDataHist("SMC","1D",RooArgList(*mass),SMCh);
cout<<"SMC hist assigned to RooDataHist"<<endl;
// Construct Pdf Model
// /\0
//gaussian
RooRealVar mean1L("mean1L","/\\ gaus 1: mean",5621.103095,5525,5700);
RooRealVar sig1L("sig1L","/\\ gaus 1: sigma",6.898126,0,100);
RooGaussian gau1L("gau1L","#Lambda signal: gaussian 1",*mass,mean1L,sig1L);
RooFormulaVar mean2L("mean2L","@0",mean1L);
RooRealVar sig2L("sig2L","/\\ gaus 2: sigma",14.693117,0,100);
RooGaussian gau2L("gau2L","#Lambda signal: gaussian 2",*mass,mean2L,sig2L);
RooRealVar f1L("f1L","/\\ signal: fraction gaussian 1",0.748776,0,1);
RooAddPdf sigL("sigL","#Lambda signal",RooArgList(gau1L,gau2L),RooArgList(f1L));
// //CB
// RooRealVar mean3L("mean3L","/\\ CB: mean",5621.001,5525,5700);
// RooRealVar sig3L("sig3L","/\\ CB: sigma",5.161,0,100);
// RooRealVar alphaL3("alphaL3","/\\ CB: alpha",2.077,0,1000);
// RooRealVar nL3("nL1","/\\ CB: n",0.286,0,1000);
// RooCBShape CBL("CBL","#Lambda signal: CB",*mass,mean3L,sig3L,alphaL3,nL3);
// RooRealVar mean4L("mean4L","/\\ gaus: mean",5621.804,5525,5700);
// RooRealVar sig4L("sig4L","/\\ gaus: sigma",10.819,0,100);
// RooGaussian gauL("gauL","#Lambda signal: gaussian",*mass,mean4L,sig4L);
// RooRealVar f1L("f1L","/\\ signal: fraction CB",0.578,0,1);
// RooAddPdf sigL("sigL","#Lambda signal",RooArgList(CBL,gauL),RooArgList(f1L));
// sigma0
//using RooHistPdf from MC--no need to build pdf here
RooHistPdf sigS = makeroohistpdf(SMC,mass,"sigS","#Sigma^{0} signal (RooHistPdf)");
// /\*
cout<<"Lst stuff"<<endl;
RooRealVar meanLst1("meanLst1","/\\*(misc.): mean1",5011.031237,4900,5100);
RooRealVar sigLst1("sigLst1","/\\*(misc.): sigma1",70.522092,0,100);
RooRealVar meanLst2("mean5Lst2","/\\*(1405): mean2",5245.261703,5100,5350);
RooRealVar sigLst2("sigLst2","/\\*(1405): sigma2",64.564763,0,100);
RooRealVar alphaLst2("alphaLst2","/\\*(1405): alpha2",29.150301);
RooRealVar nLst2("nLst2","/\\*(1405): n2",4.615817,0,50);
RooGaussian gauLst1("gauLst1","#Lambda*(misc.), gaus",*mass,meanLst1,sigLst1);
RooCBShape gauLst2("gauLst2","#Lambda*(1405), CB",*mass,meanLst2,sigLst2,alphaLst2,nLst2);
// RooRealVar fLst1("fLst1","/\\* bkg: fraction gaus 1",0.743,0,1);
// RooAddPdf bkgLst("bkgLst","#Lambda* signal",RooArgList(gauLst1,gauLst2),RooArgList(fLst1));
//Poly func BKG mass
// RooRealVar b0("b0","Background: Chebychev b0",-1.071,-10000,10000);
RooRealVar b1("b1","Background: Chebychev b1",-1.323004,-10,-0.00000000000000000000001);
RooRealVar b2("b2","Background: Chebychev b2",0.145494,0,10);
RooRealVar b3("b3","Background: Chebychev b3",-0.316,-10000,10000);
RooRealVar b4("b4","Background: Chebychev b4",0.102,-10000,10000);
RooRealVar b5("b5","Background: Chebychev b5",0.014,-10000,10000);
RooRealVar b6("b6","Background: Chebychev b6",-0.015,-10000,10000);
RooRealVar b7("b7","Background: Chebychev b7",0.012,-10000,10000);
RooArgList bList(b1,b2);
RooChebychev bkg("bkg","Background", *mass, bList);
// TF1 *ep = new TF1("ep","[2]*exp([0]*x+[1]*x*x)",4300,5950);
// ep->SetParameter(0,1);
// ep->SetParameter(1,-1);
// ep->SetParameter(2,2000);
// ep->SetParName(0,"a");
// ep->SetParName(1,"b");
// ep->SetParName(2,"c");
// RooRealVar a("a","Background: Coefficent of x",1,-10000,10000);
// RooRealVar b("b","Background: Coefficent of x*x",-1,-10000,10000);
// RooRealVar c("c","Background: Coefficent of exp()",2000,-10000,10000);
// RooTFnPdfBinding bkg("ep","ep",ep,RooArgList(*mass,a,b));
//number of each shape
RooRealVar nbkg("nbkg","N bkg",2165.490249,0,100000000);
RooRealVar nsigL("nsigL","N /\\",1689.637290,0,1000000000);
RooRealVar nsigS("nsigS","N sigma",0.000002,0,10000000000);
RooRealVar ngauLst1("ngauLst1","N /\\*(misc.)",439.812103,0,10000000000);
RooRealVar ngauLst2("ngauLst2","N /\\*(1405)",152.061617,0,10000000000);
RooRealVar nbkgLst("nbkgLst","N /\\*",591.828,0,1000000000);
//add shapes and their number to a totalPdf
RooArgList shapes;
RooArgList yields;
shapes.add(sigL); yields.add(nsigL);
shapes.add(sigS); yields.add(nsigS);
// shapes.add(bkgLst); yields.add(nbkgLst);
shapes.add(gauLst1); yields.add(ngauLst1);
shapes.add(gauLst2); yields.add(ngauLst2);
shapes.add(bkg); yields.add(nbkg);
RooAddPdf totalPdf("totalPdf","totalPdf",shapes,yields);
//fit the totalPdf
RooAbsReal * nll = totalPdf.createNLL(*data,Extended(kTRUE),Range("tot"));
RooMinuit m(*nll);
m.setVerbose(kFALSE);
m.migrad();
m.minos();
m.minos();
//display and save information
ofstream textfile;//create text file to hold data
placeholder = "plots/fit"+fileN+".txt";
textfile.open(placeholder);
TString outputtext;//for useful text
//plot things
RooPlot *framex = mass->frame();
framex->GetYaxis()->SetTitle("Events/(5 MeV)");
data->plotOn(framex,Name("Hist"),MarkerColor(kBlack),LineColor(kBlack),DataError(RooAbsData::SumW2));
totalPdf.plotOn(framex,Name("curvetot"),LineColor(kBlue));
RooArgSet* totalPdfComponents = totalPdf.getComponents();
TIterator* itertPC = totalPdfComponents->createIterator();
RooAddPdf* vartPC = (RooAddPdf*) itertPC->Next();
vartPC = (RooAddPdf*) itertPC->Next();//skip totalPdf
int i=0;//color index
TLegend *leg = new TLegend(0.2, 0.02, .4, .42);
leg->SetTextSize(0.06);
leg->AddEntry(framex->findObject("curvetot"),"Total PDF","l");
while(vartPC){//loop over compotents of totalPdf
TString vartPCtitle = vartPC->GetTitle();
TIterator* itercompPars;//forward declare so it persists outside the if statement
RooRealVar* varcompPars;
if(!(vartPCtitle.Contains(":")||vartPCtitle.Contains("@"))){//only for non-sub-shapes
while(i==0||i==10||i==4||i==1||i==5||(i>=10&&i<=27))i++;//avoid white and blue and black and yellow and horribleness
RooArgSet* compPars = vartPC->getParameters(data);//set of the parameters of the component the loop is on
itercompPars = compPars->createIterator();
varcompPars = (RooRealVar*) itercompPars->Next();
while(varcompPars){//write and print mean, sig, etc. of sub-shapes
TString vartitle = varcompPars->GetTitle();
double varval = varcompPars->getVal();
TString varvalstring = Form("%f",varval);
double hi = varcompPars->getErrorHi();
TString varerrorstring = "[exact]";
if(hi!=-1){
double lo = varcompPars->getErrorLo();
double varerror = TMath::Max(fabs(lo),hi);
varerrorstring = Form("%E",varerror);
}
outputtext = vartitle+" = "+varvalstring+" +/- "+varerrorstring;
textfile<<outputtext<<endl;
cout<<outputtext<<endl;
varcompPars = (RooRealVar*) itercompPars->Next();
}
totalPdf.plotOn(framex,Name(vartPC->GetName()),LineStyle(kDashed),LineColor(i),Components(vartPC->GetName()));
leg->AddEntry(framex->findObject(vartPC->GetName()),vartPCtitle,"l");
i++;
}
vartPC = (RooAddPdf*) itertPC->Next();
itercompPars->Reset();//make sure it's ready for the next vartPC
}
// Calculate chi2/ndf
RooArgSet *floatpar = totalPdf.getParameters(data);
int floatpars = (floatpar->selectByAttrib("Constant",kFALSE))->getSize();
Double_t chi2 = framex->chiSquare("curvetot","Hist",floatpars);
TString chi2string = Form("%f",chi2);
//create text box to list important parameters on the plot
// TPaveText* txt = new TPaveText(0.1,0.5,0.7,0.9,"NBNDC");
// txt->SetTextSize(0.06);
// txt->SetTextColor(kBlack);
// txt->SetBorderSize(0);
// txt->SetFillColor(0);
// txt->SetFillStyle(0);
outputtext = "#chi^{2}/N_{DoF} = "+chi2string;
cout<<outputtext<<endl;
textfile<<outputtext<<endl;
// txt->AddText(outputtext);
// Print stuff
TIterator* iteryields = yields.createIterator();
RooRealVar* varyields = (RooRealVar*) iteryields->Next();//only inherits things from TObject unless class specified
vector<double> Y, E;//holds yields and associated errors
vector<TString> YS, ES;//holds strings of the corresponding yields
int j=0;//count vector position
int jS=0, jL=0;//these hold the position of the S and L results;initialized in case there is no nsigS or nsigL
while(varyields){//loop over yields
TString varname = varyields->GetName();
TString vartitle = varyields->GetTitle();
double varval = varyields->getVal();
Y.push_back(varval);
double lo = varyields->getErrorLo();
double hi = varyields->getErrorHi();
E.push_back(TMath::Max(fabs(lo),hi));
YS.push_back(Form("%f",Y[j]));
ES.push_back(Form("%f",E[j]));
if(varname=="nsigS") jS=j;
if(varname=="nsigL") jL=j;
outputtext = vartitle+" = "+YS[j]+" +/- "+ES[j];
cout<<outputtext<<endl;
textfile<<outputtext<<endl;
//txt->AddText(outputtext);
varyields = (RooRealVar*) iteryields->Next();
j++;
}
//S/L
double result = Y[jS]/Y[jL];
cout<<"result declared"<<endl;
double E_result = TMath::Abs(result)*sqrt(pow(E[jS]/Y[jS],2)+pow(E[jL]/Y[jL],2));
cout<<"E_result declared"<<endl;
TString resultstring = Form("%E",result);
TString E_resultstring = Form("%E",E_result);
outputtext = "Y_{#Sigma^{0}}/Y_{#Lambda} = "+resultstring+" +/- "+E_resultstring;
cout<<outputtext<<endl;
textfile<<outputtext<<endl;
//txt->AddText(outputtext);
double resultlimit = (Y[jS]+E[jS])/(Y[jL]-E[jL]);
outputtext = Form("%E",resultlimit);
outputtext = "limit = "+outputtext;
cout<<outputtext<<endl;
textfile<<outputtext<<endl;
//txt->AddText(outputtext);
// Create canvas and pads, set style
TCanvas *c1 = new TCanvas("c1","data fits",1200,800);
TPad *pad1 = new TPad("pad1","pad1",0.0,0.3,1.0,1.0);
TPad *pad2 = new TPad("pad2","pad2",0.0,0.0,1.0,0.3);
pad1->SetBottomMargin(0);
pad2->SetTopMargin(0);
pad2->SetBottomMargin(0.5);
pad2->SetBorderMode(0);
pad1->SetBorderMode(0);
c1->SetBorderMode(0);
pad2->Draw();
pad1->Draw();
pad1->cd();
framex->SetMinimum(1);
framex->SetMaximum(3000);
framex->addObject(leg);//add legend to frame
//framex->addObject(txt);//add text to frame
gPad->SetTopMargin(0.06);
pad1->SetLogy();
// pad1->Range(4100,0,6100,0.0005);
pad1->Update();
framex->Draw();
// Pull distribution
RooPlot *framex2 = mass->frame();
RooHist* hpull = framex->pullHist("Hist","curvetot");
framex2->addPlotable(hpull,"P");
hpull->SetLineColor(kBlack);
hpull->SetMarkerColor(kBlack);
framex2->SetTitle(0);
framex2->GetYaxis()->SetTitle("Pull");
framex2->GetYaxis()->SetTitleSize(0.15);
framex2->GetYaxis()->SetLabelSize(0.15);
framex2->GetXaxis()->SetTitleSize(0.2);
framex2->GetXaxis()->SetLabelSize(0.15);
framex2->GetYaxis()->CenterTitle();
framex2->GetYaxis()->SetTitleOffset(0.45);
framex2->GetXaxis()->SetTitleOffset(1.1);
framex2->GetYaxis()->SetNdivisions(505);
framex2->GetYaxis()->SetRangeUser(-8.8,8.8);
pad2->cd();
framex2->Draw();
c1->cd();
placeholder = "plots/fit"+fileN+".eps";
c1->Print(placeholder);
placeholder = "plots/fit"+fileN+".C";
c1->SaveAs(placeholder);
textfile.close();
}
void makeAndImportDataSets(TFile *fin, RooWorkspace *wspace, RooRealVar &x){
// Get TTrees from input files and fill RooDataSets!
//const char *name = x.GetName();
RooRealVar weight("weight","weight",0,100);
weight.removeRange();
RooArgSet treevariables(x,weight);
// List all branches in the tree and add them as variables
TObjArray *branches = (TObjArray*) ((TTree*)fin->Get("data_obs"))->GetListOfBranches();
TIter next(branches);
TBranch *br;
while ( (br = (TBranch*)next()) ){
const char *brname = br->GetName();
std::cout << brname << std::endl;
if ( *brname==*weight.GetName() ) continue;
if ( *brname==*x.GetName() ) continue;
RooRealVar *vartmp = new RooRealVar(brname,brname,0,1); vartmp->removeRange();
treevariables.add(*vartmp);
}
// RooRealVar cvar(cutvar.c_str(),cutvar.c_str(),0,100);
// cvar.removeRange();
std::cout << "Building RooDataSets " << std::endl;
treevariables.Print("V");
//assert(0);
// Zvv Signal region MC sample
RooDataSet zvv("DY","DY",treevariables,RooFit::Import(*((TTree*)fin->Get("DY"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
// Zvv Control sample
// Data ... (double mu) Zvv_control
RooDataSet zvvcontrol("Zvv_control","Zvv_control",treevariables,RooFit::Import(*((TTree*)fin->Get("Zvv_control"))),RooFit::Cut(cutstring.c_str()));
// Backgrounds ... Zvv_control_bkg_mc
RooDataSet zvvcontrolbkgmc("Zvv_control_bkg_mc","Zvv_control_bkg_mc",treevariables,RooFit::Import(*((TTree*)fin->Get("T_control_bkg_mc"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet zvvcontrolbkgmc_1("1","1",treevariables,RooFit::Import(*(TTree*)fin->Get("TT_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet zvvcontrolbkgmc_2("2","2",treevariables,RooFit::Import(*(TTree*)fin->Get("WW_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet zvvcontrolbkgmc_3("3","3",treevariables,RooFit::Import(*(TTree*)fin->Get("WZ_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet zvvcontrolbkgmc_4("4","4",treevariables,RooFit::Import(*(TTree*)fin->Get("ZZ_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
zvvcontrolbkgmc.append(zvvcontrolbkgmc_1);
zvvcontrolbkgmc.append(zvvcontrolbkgmc_2);
zvvcontrolbkgmc.append(zvvcontrolbkgmc_3);
zvvcontrolbkgmc.append(zvvcontrolbkgmc_4);
// MC ... Zvv_control_mc
RooDataSet zvvcontrolmc("Zvv_control_mc","Zvv_control_mc",treevariables,RooFit::Import(*((TTree*)fin->Get("Zvv_control_mc"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
std::cout << " Weighs of datasets " << std::endl;
std::cout << " Zvv_control " << zvvcontrol.sumEntries() << std::endl;
std::cout << " Zvv_control_mc " << zvvcontrolmc.sumEntries() << std::endl;
// Wlv Control sample (single mu)
// Data ... (double mu)
// Backgrounds ...
//.append()
// MC ...
// Store these to the output workspace
wspace->import(zvvcontrol);
wspace->import(zvvcontrolmc);
wspace->import(zvvcontrolbkgmc);
wspace->import(zvv);
// Wlv Signal region MC sample
RooDataSet wlv("W","W",treevariables,RooFit::Import(*((TTree*)fin->Get("W"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlv_ht("WHT","WHT",treevariables,RooFit::Import(*((TTree*)fin->Get("WHT"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
wlv.append(wlv_ht);
// Wlv Control sample
// Data ... (single mu) Wlv_control
RooDataSet wlvcontrol("Wlv_control","Wlv_control",treevariables,RooFit::Import(*((TTree*)fin->Get("Wlv_control"))),RooFit::Cut(cutstring.c_str()));
// Backgrounds ... Wlv_control_bkg_mc
RooDataSet wlvcontrolbkgmc("Wlv_control_bkg_mc","Wlv_control_bkg_mc",treevariables,RooFit::Import(*((TTree*)fin->Get("T_sl_control_bkg_mc"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolbkgmc_1("1","1",treevariables,RooFit::Import(*(TTree*)fin->Get("TT_sl_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolbkgmc_2("2","2",treevariables,RooFit::Import(*(TTree*)fin->Get("DY_sl_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolbkgmc_3("3","3",treevariables,RooFit::Import(*(TTree*)fin->Get("WZ_sl_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolbkgmc_4("4","4",treevariables,RooFit::Import(*(TTree*)fin->Get("ZZ_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolbkgmc_5("5","5",treevariables,RooFit::Import(*(TTree*)fin->Get("WW_control_bkg_mc")),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
wlvcontrolbkgmc.append(wlvcontrolbkgmc_1);
wlvcontrolbkgmc.append(wlvcontrolbkgmc_2);
wlvcontrolbkgmc.append(wlvcontrolbkgmc_3);
wlvcontrolbkgmc.append(wlvcontrolbkgmc_4);
wlvcontrolbkgmc.append(wlvcontrolbkgmc_5);
// MC ... Zvv_control_mc
RooDataSet wlvcontrolmc("Wlv_control_mc","Wlv_control_mc",treevariables,RooFit::Import(*((TTree*)fin->Get("Wlv_control_mc_1"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
RooDataSet wlvcontrolmc_2("Wlv_control_mc_2","Wlv_control_mc_2",treevariables,RooFit::Import(*((TTree*)fin->Get("Wlv_control_mc_2"))),RooFit::Cut(cutstring.c_str()),RooFit::WeightVar("weight"));
wlvcontrolmc.append(wlvcontrolmc_2);
std::cout << " Weighs of datasets " << std::endl;
std::cout << " Wlv_control " << wlvcontrol.sumEntries() << std::endl;
std::cout << " Wlv_control_mc " << wlvcontrolmc.sumEntries() << std::endl;
// Wlv Control sample (single mu)
// Data ... (double mu)
// Backgrounds ...
//.append()
// MC ...
// Store these to the output workspace
wspace->import(wlvcontrol);
wspace->import(wlvcontrolmc);
wspace->import(wlvcontrolbkgmc);
wspace->import(wlv);
}
void StandardBayesianNumericalDemo(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
/////////////////////////////////////////////////////////////
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
////////////////////////////////////////////////////////////
TString filename = infile;
if (filename.IsNull()) {
filename = "results/example_combined_GaussExample_model.root";
std::cout << "Use standard file generated with HistFactory : " << filename << std::endl;
}
// Check if example input file exists
TFile *file = TFile::Open(filename);
// if input file was specified but not found, quit
if(!file && !TString(infile).IsNull()){
cout <<"file " << filename << " not found" << endl;
return;
}
// if default file not found, try to create it
if(!file ){
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
// now try to access the file again
file = TFile::Open(filename);
}
if(!file){
// if it is still not there, then we can't continue
cout << "Not able to run hist2workspace to create example input" <<endl;
return;
}
/////////////////////////////////////////////////////////////
// Tutorial starts here
////////////////////////////////////////////////////////////
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
/////////////////////////////////////////////
// create and use the BayesianCalculator
// to find and plot the 95% credible interval
// on the parameter of interest as specified
// in the model config
// before we do that, we must specify our prior
// it belongs in the model config, but it may not have
// been specified
RooUniform prior("prior","",*mc->GetParametersOfInterest());
w->import(prior);
mc->SetPriorPdf(*w->pdf("prior"));
// do without systematics
//mc->SetNuisanceParameters(RooArgSet() );
BayesianCalculator bayesianCalc(*data,*mc);
bayesianCalc.SetConfidenceLevel(0.95); // 95% interval
// default of the calculator is central interval. here use shortest , central or upper limit depending on input
// doing a shortest interval might require a longer time since it requires a scan of the posterior function
if (intervalType == 0) bayesianCalc.SetShortestInterval(); // for shortest interval
if (intervalType == 1) bayesianCalc.SetLeftSideTailFraction(0.5); // for central interval
if (intervalType == 2) bayesianCalc.SetLeftSideTailFraction(0.); // for upper limit
if (!integrationType.IsNull() ) {
bayesianCalc.SetIntegrationType(integrationType); // set integrationType
bayesianCalc.SetNumIters(nToys); // set number of ietrations (i.e. number of toys for MC integrations)
}
// compute interval by scanning the posterior function
if (scanPosterior)
bayesianCalc.SetScanOfPosterior(nScanPoints);
SimpleInterval* interval = bayesianCalc.GetInterval();
// print out the iterval on the first Parameter of Interest
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
cout << "\n95% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit() << ", "<<
interval->UpperLimit() <<"] "<<endl;
// make a plot
// since plotting may take a long time (it requires evaluating
// the posterior in many points) this command will speed up
// by reducing the number of points to plot - do 50
cout << "\nDrawing plot of posterior function....." << endl;
bayesianCalc.SetScanOfPosterior(nScanPoints);
RooPlot * plot = bayesianCalc.GetPosteriorPlot();
plot->Draw();
}
void StandardTestStatDistributionDemo(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
// the number of toy MC used to generate the distribution
int nToyMC = 1000;
// The parameter below is needed for asymptotic distribution to be chi-square,
// but set to false if your model is not numerically stable if mu<0
bool allowNegativeMu=true;
/////////////////////////////////////////////////////////////
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
////////////////////////////////////////////////////////////
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
/////////////////////////////////////////////////////////////
// Now get the data and workspace
////////////////////////////////////////////////////////////
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
mc->Print();
/////////////////////////////////////////////////////////////
// Now find the upper limit based on the asymptotic results
////////////////////////////////////////////////////////////
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
ProfileLikelihoodCalculator plc(*data,*mc);
LikelihoodInterval* interval = plc.GetInterval();
double plcUpperLimit = interval->UpperLimit(*firstPOI);
delete interval;
cout << "\n\n--------------------------------------"<<endl;
cout <<"Will generate sampling distribution at " << firstPOI->GetName() << " = " << plcUpperLimit <<endl;
int nPOI = mc->GetParametersOfInterest()->getSize();
if(nPOI>1){
cout <<"not sure what to do with other parameters of interest, but here are their values"<<endl;
mc->GetParametersOfInterest()->Print("v");
}
/////////////////////////////////////////////
// create thte test stat sampler
ProfileLikelihoodTestStat ts(*mc->GetPdf());
// to avoid effects from boundary and simplify asymptotic comparison, set min=-max
if(allowNegativeMu)
firstPOI->setMin(-1*firstPOI->getMax());
// temporary RooArgSet
RooArgSet poi;
poi.add(*mc->GetParametersOfInterest());
// create and configure the ToyMCSampler
ToyMCSampler sampler(ts,nToyMC);
sampler.SetPdf(*mc->GetPdf());
sampler.SetObservables(*mc->GetObservables());
sampler.SetGlobalObservables(*mc->GetGlobalObservables());
if(!mc->GetPdf()->canBeExtended() && (data->numEntries()==1)){
cout << "tell it to use 1 event"<<endl;
sampler.SetNEventsPerToy(1);
}
firstPOI->setVal(plcUpperLimit); // set POI value for generation
sampler.SetParametersForTestStat(*mc->GetParametersOfInterest()); // set POI value for evaluation
if (useProof) {
ProofConfig pc(*w, nworkers, "",false);
sampler.SetProofConfig(&pc); // enable proof
}
firstPOI->setVal(plcUpperLimit);
RooArgSet allParameters;
allParameters.add(*mc->GetParametersOfInterest());
allParameters.add(*mc->GetNuisanceParameters());
allParameters.Print("v");
SamplingDistribution* sampDist = sampler.GetSamplingDistribution(allParameters);
SamplingDistPlot plot;
plot.AddSamplingDistribution(sampDist);
plot.GetTH1F(sampDist)->GetYaxis()->SetTitle(Form("f(-log #lambda(#mu=%.2f) | #mu=%.2f)",plcUpperLimit,plcUpperLimit));
plot.SetAxisTitle(Form("-log #lambda(#mu=%.2f)",plcUpperLimit));
TCanvas* c1 = new TCanvas("c1");
c1->SetLogy();
plot.Draw();
double min = plot.GetTH1F(sampDist)->GetXaxis()->GetXmin();
double max = plot.GetTH1F(sampDist)->GetXaxis()->GetXmax();
TF1* f = new TF1("f",Form("2*ROOT::Math::chisquared_pdf(2*x,%d,0)",nPOI),min,max);
f->Draw("same");
c1->SaveAs("standard_test_stat_distribution.pdf");
}
void CountingModelCross_2( int nobs_4m = 27, // number of observed events
double b_4m = 0.09, // number of background events
double sigmab_4m = 0.16, // relative uncertainty in b 1
int nobs_2e2m = 30, // number of observed events
double b_2e2m = 0.47, // number of background events
double sigmab_2e2m = 0.26, // relative uncertainty in b 2
int nobs_4e = 8, // number of observed events
double b_4e = 0.44, // number of background events
double sigmab_4e = 0.35 // relative uncertainty in b 2
)
{
RooWorkspace w("w");
// make Poisson model * Gaussian constraint
w.factory("prod:s_4m(cross[34.7,0.,10000000],BR_4m[0.5,0,1],Acc_4m[0.5,0.,1.],Lumi[2.229,0.,10000])");
w.factory("prod:s_2e2m(cross[34.7,0.,10000000],BR_2e2m[0.5,0,1],Acc_2e2m[0.5,0.,1.],Lumi[2.229,0.,10000])");
w.factory("prod:s_4e(cross[34.7,0.,10000000],BR_4e[0.5,0,1],Acc_4e[0.5,0.,1.],Lumi[2.229,0.,10000])");
w.factory("sum:nexp_4m(s_4m,b_4m[1,0,10])");
w.factory("sum:nexp_2e2m(s_2e2m,b_2e2m[1,0,10])");
w.factory("sum:nexp_4e(s_4e,b_4e[1,0,10])");
w.var("b_4m")->setVal(b_4m);
w.var("b_2e2m")->setVal(b_2e2m);
w.var("b_4e")->setVal(b_4e);
// Poisson of (n | s+b)
w.factory("Poisson:pdf_4m(nobs_4m[0,50],nexp_4m)");
w.factory("Poisson:pdf_2e2m(nobs_2e2m[0,50],nexp_2e2m)");
w.factory("Poisson:pdf_4e(nobs_4e[0,50],nexp_4e)");
w.factory("Gaussian:constraint_4m(b0_4m[0,10],b_4m,sigmab_4m[1])");
w.factory("Gaussian:constraint_2e2m(b0_2e2m[0,10],b_2e2m,sigmab_2e2m[1])");
w.factory("Gaussian:constraint_4e(b0_4e[0,10],b_4e,sigmab_4e[1])");
w.factory("Gaussian:constraintAcc_4m(Acc0_4m[0,1],Acc_4m,sigmaAcc_4m[1])");
w.factory("Gaussian:constraintAcc_2e2m(Acc0_2e2m[0,1],Acc_2e2m,sigmaAcc_2e2m[1])");
w.factory("Gaussian:constraintAcc_4e(Acc0_4e[0,1],Acc_4e,sigmaAcc_4e[1])");
w.factory("PROD:model_4m(pdf_4m,constraint_4m,constraintAcc_4m)");
w.factory("PROD:model_2e2m(pdf_2e2m,constraint_2e2m,constraintAcc_2e2m)");
w.factory("PROD:model_4e(pdf_4e,constraint_4e,constraintAcc_4e)");
w.factory("PROD:model(model_4m,model_2e2m,model_4e)");
w.var("b0_4m")->setVal(b_4m);
w.var("b0_2e2m")->setVal(b_2e2m);
w.var("b0_4e")->setVal(b_4e);
w.var("b0_4m")->setConstant(true); // needed for being treated as global observables
w.var("b0_2e2m")->setConstant(true); // needed for being treated as global observables
w.var("b0_4e")->setConstant(true); // needed for being treated as global observables
// w.var("Acc_4m")->setVal(0.844);
// w.var("Acc_2e2m")->setVal(0.694);
// w.var("Acc_4e")->setVal(0.60);
Float_t Acc_4m = 0.844*0.5388;
Float_t Acc_2e2m = 0.694*0.5388;
Float_t Acc_4e = 0.60*0.5388;
// Float_t Acc_4m = 0.844;
// Float_t Acc_2e2m = 0.694;
// Float_t Acc_4e = 0.60;
Float_t sigmaAcc_4m = 0.0097 ;
Float_t sigmaAcc_2e2m = 0.0097 ;
Float_t sigmaAcc_4e = 0.0097 ;
Float_t BRele = 0.03363;
Float_t BRmu = 0.03366;
Float_t BR_4m = BRmu*BRmu;
Float_t BR_2e2m = 2*BRmu*BRele;
Float_t BR_4e = BRele*BRele;
Float_t BR = BR_4m+BR_2e2m+BR_4e;
cout<<"BR sum"<<BR<<" BRmean "<<(BR_4m + BR_4e + BR_2e2m)/3.<<" BR_4m "<<BR_4m<<" BR_2e2m "<<BR_2e2m<<" BR_4e "<<BR_4e<<endl;
w.var("BR_4m")->setVal(BR_4m);
w.var("BR_2e2m")->setVal(BR_2e2m);
w.var("BR_4e")->setVal(BR_4e);
w.var("BR_4m")->setConstant(true); // needed for being treated as global observables
w.var("BR_2e2m")->setConstant(true); // needed for being treated as global observables
w.var("BR_4e")->setConstant(true); // needed for being treated as global observables
w.var("Lumi")->setVal(2.560);
w.var("Lumi")->setConstant(true);
w.var("sigmab_4m")->setVal(sigmab_4m*b_4m);
w.var("sigmab_2e2m")->setVal(sigmab_2e2m*b_2e2m);
w.var("sigmab_4e")->setVal(sigmab_4e*b_4e);
w.var("sigmaAcc_4m")->setVal(sigmaAcc_4m*Acc_4m);
w.var("sigmaAcc_2e2m")->setVal(sigmaAcc_2e2m*Acc_2e2m);
w.var("sigmaAcc_4e")->setVal(sigmaAcc_4e*Acc_4e);
ModelConfig mc("ModelConfig",&w);
mc.SetPdf(*w.pdf("model"));
mc.SetParametersOfInterest(*w.var("cross"));
mc.SetObservables(RooArgSet(*w.var("nobs_4m"),*w.var("nobs_2e2m"),*w.var("nobs_4e")));
// mc.SetNuisanceParameters(RooArgSet(*w.var("b_4m"),*w.var("b_2e2m"),*w.var("b_4e")));
//mc.SetObservables(RooArgSet(*w.var("nobs_4m"),*w.var("nobs_2e2m")));
// mc.SetNuisanceParameters(RooArgSet(*w.var("b_4m"),*w.var("b_2e2m"),*w.var("Acc_4m"),*w.var("Acc_2e2m")));
mc.SetNuisanceParameters(RooArgSet(*w.var("b_4m"),*w.var("b_2e2m"),*w.var("b_4e"),*w.var("Acc_4m"),*w.var("Acc_2e2m"),*w.var("Acc_4e")));
mc.SetParametersOfInterest(*w.var("cross"));
// these are needed for the hypothesis tests
mc.SetSnapshot(*w.var("cross"));
// mc.SetGlobalObservables(RooArgSet(*w.var("b0_4m"),*w.var("b0_2e2m"),*w.var("b0_4e")));
mc.Print();
// import model in the workspace
w.import(mc);
// make data set with the namber of observed events
RooDataSet data("data","", RooArgSet(*w.var("nobs_4m"),*w.var("nobs_2e2m"),*w.var("nobs_4e")));
//RooDataSet data("data","", RooArgSet(*w.var("nobs_4m"),*w.var("nobs_2e2m")));
w.var("nobs_4m")->setVal(nobs_4m);
data.add(*w.var("nobs_4m") );
w.var("nobs_2e2m")->setVal(nobs_2e2m);
data.add(*w.var("nobs_2e2m") );
w.var("nobs_4e")->setVal(nobs_4e);
data.add(*w.var("nobs_4e") );
// import data set in workspace and save it in a file
ProfileLikelihoodCalculator pl(data, mc);
pl.SetConfidenceLevel(0.68); // 95% interval
LikelihoodInterval* interval = pl.GetInterval();
// print out the iterval on the first Parameter of Interest
RooRealVar* firstPOI = (RooRealVar*) mc.GetParametersOfInterest()->first();
cout << "\n65% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit(*firstPOI) << ", "<<
interval->UpperLimit(*firstPOI) <<"] "<<endl;
std::cout<<"Cross "<<mc.GetParametersOfInterest()->getRealValue("cross")*BR<<std::endl;
w.import(data);
w.Print();
TString fileName = "CountingModel.root";
// write workspace in the file (recreate file if already existing)
w.writeToFile(fileName, true);
}
///
/// Print this PDF in a verbose way:
/// - observables
/// - correlations
/// - parameters
///
void PDF_Abs::print() const
{
TString cleanName = name;
cout << "PDF: " << cleanName.ReplaceAll(uniqueID,"") << " (" << uniqueID << ")" << endl << endl;
if ( observables ){
cout << " observables:" << endl;
cout << " nObs = " << nObs << endl;
cout << " values from: " << obsValSource << endl;
cout << " errors from: " << obsErrSource << endl;
for ( int iObs=0; iObs<nObs; iObs++ ){
RooRealVar* v = (RooRealVar*)observables->at(iObs);
TString obsName = v->GetName();
obsName.ReplaceAll(uniqueID,"");
printf(" %-20s = %8.5f +/- %7.5f +/- %7.5f\n", obsName.Data(), v->getVal(), StatErr[iObs], SystErr[iObs]);
}
}
else cout << "PDF_Abs::print() : observables not initialized. Call initObservables() first." << endl;
cout << endl;
if ( nObs>1 ){
printCorMatrix("(stat+syst)", corSource, corMatrix);
printCorMatrix("(stat)", corSource, corStatMatrix);
printCorMatrix("(syst)", corSource, corSystMatrix);
}
if ( parameters ){
cout << " parameters:" << endl;
cout << " nPar = " << parameters->getSize() << endl;
cout << " ";
bool first=true;
TIterator* it = parameters->createIterator();
while ( RooAbsReal* v = (RooAbsReal*)it->Next() ){
cout << (first?"":", ") << v->GetName();
first=false;
}
delete it;
cout << endl;
}
else cout << "PDF_Abs::print() : parameters not initialized. Call initParameters() first." << endl;
cout << endl;
if ( theory ){
cout << " relations:" << endl;
TIterator* it = theory->createIterator();
while ( RooAbsReal* v = (RooAbsReal*)it->Next() ){
// it's not easy to extract the formula from a RooFormulaVar.
TString thName = v->GetName();
thName.ReplaceAll(uniqueID,"");
printf(" %-20s = ", thName.Data());
ostringstream stream;
v->printMetaArgs(stream);
TString formula = stream.str();
formula.ReplaceAll("formula=", "");
formula.ReplaceAll("\"", "");
if ( formula=="" ) formula = v->ClassName(); // compiled custom Roo*Var classes don't have a formula
cout << formula << endl;
}
delete it;
}
else cout << "PDF_Abs::print() : theory not initialized. Call initRelations() first." << endl;
cout << endl;
}
void PDF_GLWADS_DKDpi_K3pi::setUncertainties(config c)
{
switch(c)
{
case lumi1fb:{
obsErrSource = "1fb-1, ExpNll/sept2012K3PIResult.root";
TString File = this->dir+"/ExpNll/sept2012K3PIResult.root";
TFile *fr = TFile::Open(File);
RooFitResult *r = (RooFitResult*)fr->Get("fitresult_model_reducedData_binned");
assert(r);
for ( int i=0; i<nObs; i++ )
{
RooRealVar* pObs = (RooRealVar*)((RooArgList*)observables)->at(i);
RooRealVar* pRes = (RooRealVar*)r->floatParsFinal().find(obsTmkToMalcolm(pObs->GetName()));
assert(pRes);
StatErr[i] = pRes->getError();
}
SystErr[0] = 0.0026; // rkp_k3pi_obs
SystErr[1] = 0.018; // afav_dk_obs
SystErr[2] = 0.010; // afav_dpi_obs
SystErr[3] = 0.0010; // rp_dk_obs
SystErr[4] = 0.0008; // rm_dk_obs
SystErr[5] = 0.00011; // rp_dpi_obs
SystErr[6] = 0.00011; // rm_dpi_obs
fr->Close();
delete r;
delete fr;
break;
}
case lumi3fb:{
obsErrSource = "3fb-1 ANA v7 unblind"; // https://twiki.cern.ch/twiki/pub/LHCbPhysics/B2D0K/LHCb-ANA-2014-071-v7.pdf (see Vavas email 04/08/15)
// these get transformed over from the new inputs using ExpNll/transportGLWADS_new_to_old.py
// in the case of the DK only (robust) combination some of the observables don't exist
// usemap as the temp store
std::map< TString, double > stat_errs;
stat_errs["rkp_k3pi_obs"] = 0.0010; // rkp_k3pi_obs
stat_errs["afav_dk_k3pi_obs"] = 0.0119; // afav_dk_obs
stat_errs["afav_dpi_k3pi_obs"] = 0.1; // afav_dpi_obs // now Aprod, large error to disable it
stat_errs["rp_dk_k3pi_obs"] = 0.002421; // rp_dk_obs
stat_errs["rm_dk_k3pi_obs"] = 0.001756; // rm_dk_obs
stat_errs["rp_dpi_k3pi_obs"] = 0.000252; // rp_dpi_obs
stat_errs["rm_dpi_k3pi_obs"] = 0.000258; // rm_dpi_obs
for ( int i=0; i<nObs; i++ )
{
RooRealVar* pObs = (RooRealVar*)((RooArgList*)observables)->at(i);
StatErr[i] = stat_errs[pObs->GetName()];
}
stat_errs.clear();
std::map< TString, double > syst_errs;
syst_errs["rkp_k3pi_obs"] = 0.0170; // rkp_k3pi_obs
syst_errs["afav_dk_k3pi_obs"] = 0.0019; // afav_dk_obs
syst_errs["afav_dpi_k3pi_obs"] = 0.1; // afav_dpi_obs // now Aprod, large error to disable it
syst_errs["rp_dk_k3pi_obs"] = 0.000950; // rp_dk_obs
syst_errs["rm_dk_k3pi_obs"] = 0.000673; // rm_dk_obs
syst_errs["rp_dpi_k3pi_obs"] = 0.000061; // rp_dpi_obs
syst_errs["rm_dpi_k3pi_obs"] = 0.000064; // rm_dpi_obs
for ( int i=0; i<nObs; i++ )
{
RooRealVar* pObs = (RooRealVar*)((RooArgList*)observables)->at(i);
SystErr[i] = syst_errs[pObs->GetName()];
}
syst_errs.clear();
break;
}
case lumi9fb:{
setUncertainties(lumi3fb);
obsErrSource = "9fb-1 errors obtained from scaling down the 3fb-1 errors (lumi3fb)";
for ( int i=0; i<nObs; i++ ){
StatErr[i] /= (sqrt(2.7));
SystErr[i] /= (sqrt(2.7));
}
break;
}
default:{
cout << "PDF_GLWADS_DKDpi_K3pi::setUncertainties() : ERROR : config "+ConfigToTString(c)+" not found." << endl;
exit(1);
}
}
}
void OneSidedFrequentistUpperLimitWithBands(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData") {
double confidenceLevel=0.95;
int nPointsToScan = 20;
int nToyMC = 200;
// -------------------------------------------------------
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
// -------------------------------------------------------
// Now get the data and workspace
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
// -------------------------------------------------------
// Now get the POI for convenience
// you may want to adjust the range of your POI
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
/* firstPOI->setMin(0);*/
/* firstPOI->setMax(10);*/
// --------------------------------------------
// Create and use the FeldmanCousins tool
// to find and plot the 95% confidence interval
// on the parameter of interest as specified
// in the model config
// REMEMBER, we will change the test statistic
// so this is NOT a Feldman-Cousins interval
FeldmanCousins fc(*data,*mc);
fc.SetConfidenceLevel(confidenceLevel);
/* fc.AdditionalNToysFactor(0.25); // degrade/improve sampling that defines confidence belt*/
/* fc.UseAdaptiveSampling(true); // speed it up a bit, don't use for expected limits*/
fc.SetNBins(nPointsToScan); // set how many points per parameter of interest to scan
fc.CreateConfBelt(true); // save the information in the belt for plotting
// -------------------------------------------------------
// Feldman-Cousins is a unified limit by definition
// but the tool takes care of a few things for us like which values
// of the nuisance parameters should be used to generate toys.
// so let's just change the test statistic and realize this is
// no longer "Feldman-Cousins" but is a fully frequentist Neyman-Construction.
/* ProfileLikelihoodTestStatModified onesided(*mc->GetPdf());*/
/* fc.GetTestStatSampler()->SetTestStatistic(&onesided);*/
/* ((ToyMCSampler*) fc.GetTestStatSampler())->SetGenerateBinned(true); */
ToyMCSampler* toymcsampler = (ToyMCSampler*) fc.GetTestStatSampler();
ProfileLikelihoodTestStat* testStat = dynamic_cast<ProfileLikelihoodTestStat*>(toymcsampler->GetTestStatistic());
testStat->SetOneSided(true);
// Since this tool needs to throw toy MC the PDF needs to be
// extended or the tool needs to know how many entries in a dataset
// per pseudo experiment.
// In the 'number counting form' where the entries in the dataset
// are counts, and not values of discriminating variables, the
// datasets typically only have one entry and the PDF is not
// extended.
if(!mc->GetPdf()->canBeExtended()){
if(data->numEntries()==1)
fc.FluctuateNumDataEntries(false);
else
cout <<"Not sure what to do about this model" <<endl;
}
// We can use PROOF to speed things along in parallel
// However, the test statistic has to be installed on the workers
// so either turn off PROOF or include the modified test statistic
// in your `$ROOTSYS/roofit/roostats/inc` directory,
// add the additional line to the LinkDef.h file,
// and recompile root.
if (useProof) {
ProofConfig pc(*w, nworkers, "", false);
toymcsampler->SetProofConfig(&pc); // enable proof
}
if(mc->GetGlobalObservables()){
cout << "will use global observables for unconditional ensemble"<<endl;
mc->GetGlobalObservables()->Print();
toymcsampler->SetGlobalObservables(*mc->GetGlobalObservables());
}
// Now get the interval
PointSetInterval* interval = fc.GetInterval();
ConfidenceBelt* belt = fc.GetConfidenceBelt();
// print out the interval on the first Parameter of Interest
cout << "\n95% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit(*firstPOI) << ", "<<
interval->UpperLimit(*firstPOI) <<"] "<<endl;
// get observed UL and value of test statistic evaluated there
RooArgSet tmpPOI(*firstPOI);
double observedUL = interval->UpperLimit(*firstPOI);
firstPOI->setVal(observedUL);
double obsTSatObsUL = fc.GetTestStatSampler()->EvaluateTestStatistic(*data,tmpPOI);
// Ask the calculator which points were scanned
RooDataSet* parameterScan = (RooDataSet*) fc.GetPointsToScan();
RooArgSet* tmpPoint;
// make a histogram of parameter vs. threshold
TH1F* histOfThresholds = new TH1F("histOfThresholds","",
parameterScan->numEntries(),
firstPOI->getMin(),
firstPOI->getMax());
histOfThresholds->GetXaxis()->SetTitle(firstPOI->GetName());
histOfThresholds->GetYaxis()->SetTitle("Threshold");
// loop through the points that were tested and ask confidence belt
// what the upper/lower thresholds were.
// For FeldmanCousins, the lower cut off is always 0
for(Int_t i=0; i<parameterScan->numEntries(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
//cout <<"get threshold"<<endl;
double arMax = belt->GetAcceptanceRegionMax(*tmpPoint);
double poiVal = tmpPoint->getRealValue(firstPOI->GetName()) ;
histOfThresholds->Fill(poiVal,arMax);
}
TCanvas* c1 = new TCanvas();
c1->Divide(2);
c1->cd(1);
histOfThresholds->SetMinimum(0);
histOfThresholds->Draw();
c1->cd(2);
// -------------------------------------------------------
// Now we generate the expected bands and power-constraint
// First: find parameter point for mu=0, with conditional MLEs for nuisance parameters
RooAbsReal* nll = mc->GetPdf()->createNLL(*data);
RooAbsReal* profile = nll->createProfile(*mc->GetParametersOfInterest());
firstPOI->setVal(0.);
profile->getVal(); // this will do fit and set nuisance parameters to profiled values
RooArgSet* poiAndNuisance = new RooArgSet();
if(mc->GetNuisanceParameters())
poiAndNuisance->add(*mc->GetNuisanceParameters());
poiAndNuisance->add(*mc->GetParametersOfInterest());
w->saveSnapshot("paramsToGenerateData",*poiAndNuisance);
RooArgSet* paramsToGenerateData = (RooArgSet*) poiAndNuisance->snapshot();
cout << "\nWill use these parameter points to generate pseudo data for bkg only" << endl;
paramsToGenerateData->Print("v");
RooArgSet unconditionalObs;
unconditionalObs.add(*mc->GetObservables());
unconditionalObs.add(*mc->GetGlobalObservables()); // comment this out for the original conditional ensemble
double CLb=0;
double CLbinclusive=0;
// Now we generate background only and find distribution of upper limits
TH1F* histOfUL = new TH1F("histOfUL","",100,0,firstPOI->getMax());
histOfUL->GetXaxis()->SetTitle("Upper Limit (background only)");
histOfUL->GetYaxis()->SetTitle("Entries");
for(int imc=0; imc<nToyMC; ++imc){
// set parameters back to values for generating pseudo data
// cout << "\n get current nuis, set vals, print again" << endl;
w->loadSnapshot("paramsToGenerateData");
// poiAndNuisance->Print("v");
RooDataSet* toyData = 0;
// now generate a toy dataset
if(!mc->GetPdf()->canBeExtended()){
if(data->numEntries()==1)
toyData = mc->GetPdf()->generate(*mc->GetObservables(),1);
else
cout <<"Not sure what to do about this model" <<endl;
} else{
// cout << "generating extended dataset"<<endl;
toyData = mc->GetPdf()->generate(*mc->GetObservables(),Extended());
}
// generate global observables
// need to be careful for simpdf
// RooDataSet* globalData = mc->GetPdf()->generate(*mc->GetGlobalObservables(),1);
RooSimultaneous* simPdf = dynamic_cast<RooSimultaneous*>(mc->GetPdf());
if(!simPdf){
RooDataSet *one = mc->GetPdf()->generate(*mc->GetGlobalObservables(), 1);
const RooArgSet *values = one->get();
RooArgSet *allVars = mc->GetPdf()->getVariables();
*allVars = *values;
delete allVars;
delete values;
delete one;
} else {
//try fix for sim pdf
TIterator* iter = simPdf->indexCat().typeIterator() ;
RooCatType* tt = NULL;
while((tt=(RooCatType*) iter->Next())) {
// Get pdf associated with state from simpdf
RooAbsPdf* pdftmp = simPdf->getPdf(tt->GetName()) ;
// Generate only global variables defined by the pdf associated with this state
RooArgSet* globtmp = pdftmp->getObservables(*mc->GetGlobalObservables()) ;
RooDataSet* tmp = pdftmp->generate(*globtmp,1) ;
// Transfer values to output placeholder
*globtmp = *tmp->get(0) ;
// Cleanup
delete globtmp ;
delete tmp ;
}
}
// globalData->Print("v");
// unconditionalObs = *globalData->get();
// mc->GetGlobalObservables()->Print("v");
// delete globalData;
// cout << "toy data = " << endl;
// toyData->get()->Print("v");
// get test stat at observed UL in observed data
firstPOI->setVal(observedUL);
double toyTSatObsUL = fc.GetTestStatSampler()->EvaluateTestStatistic(*toyData,tmpPOI);
// toyData->get()->Print("v");
// cout <<"obsTSatObsUL " <<obsTSatObsUL << "toyTS " << toyTSatObsUL << endl;
if(obsTSatObsUL < toyTSatObsUL) // not sure about <= part yet
CLb+= (1.)/nToyMC;
if(obsTSatObsUL <= toyTSatObsUL) // not sure about <= part yet
CLbinclusive+= (1.)/nToyMC;
// loop over points in belt to find upper limit for this toy data
double thisUL = 0;
for(Int_t i=0; i<parameterScan->numEntries(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
double arMax = belt->GetAcceptanceRegionMax(*tmpPoint);
firstPOI->setVal( tmpPoint->getRealValue(firstPOI->GetName()) );
// double thisTS = profile->getVal();
double thisTS = fc.GetTestStatSampler()->EvaluateTestStatistic(*toyData,tmpPOI);
// cout << "poi = " << firstPOI->getVal()
// << " max is " << arMax << " this profile = " << thisTS << endl;
// cout << "thisTS = " << thisTS<<endl;
if(thisTS<=arMax){
thisUL = firstPOI->getVal();
} else{
break;
}
}
/*
// loop over points in belt to find upper limit for this toy data
double thisUL = 0;
for(Int_t i=0; i<histOfThresholds->GetNbinsX(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
cout <<"---------------- "<<i<<endl;
tmpPoint->Print("v");
cout << "from hist " << histOfThresholds->GetBinCenter(i+1) <<endl;
double arMax = histOfThresholds->GetBinContent(i+1);
// cout << " threhold from Hist = aMax " << arMax<<endl;
// double arMax2 = belt->GetAcceptanceRegionMax(*tmpPoint);
// cout << "from scan arMax2 = "<< arMax2 << endl; // not the same due to TH1F not TH1D
// cout << "scan - hist" << arMax2-arMax << endl;
firstPOI->setVal( histOfThresholds->GetBinCenter(i+1));
// double thisTS = profile->getVal();
double thisTS = fc.GetTestStatSampler()->EvaluateTestStatistic(*toyData,tmpPOI);
// cout << "poi = " << firstPOI->getVal()
// << " max is " << arMax << " this profile = " << thisTS << endl;
// cout << "thisTS = " << thisTS<<endl;
// NOTE: need to add a small epsilon term for single precision vs. double precision
if(thisTS<=arMax + 1e-7){
thisUL = firstPOI->getVal();
} else{
break;
}
}
*/
histOfUL->Fill(thisUL);
// for few events, data is often the same, and UL is often the same
// cout << "thisUL = " << thisUL<<endl;
delete toyData;
}
histOfUL->Draw();
c1->SaveAs("one-sided_upper_limit_output.pdf");
// if you want to see a plot of the sampling distribution for a particular scan point:
/*
SamplingDistPlot sampPlot;
int indexInScan = 0;
tmpPoint = (RooArgSet*) parameterScan->get(indexInScan)->clone("temp");
firstPOI->setVal( tmpPoint->getRealValue(firstPOI->GetName()) );
toymcsampler->SetParametersForTestStat(tmpPOI);
SamplingDistribution* samp = toymcsampler->GetSamplingDistribution(*tmpPoint);
sampPlot.AddSamplingDistribution(samp);
sampPlot.Draw();
*/
// Now find bands and power constraint
Double_t* bins = histOfUL->GetIntegral();
TH1F* cumulative = (TH1F*) histOfUL->Clone("cumulative");
cumulative->SetContent(bins);
double band2sigDown, band1sigDown, bandMedian, band1sigUp,band2sigUp;
for(int i=1; i<=cumulative->GetNbinsX(); ++i){
if(bins[i]<RooStats::SignificanceToPValue(2))
band2sigDown=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(1))
band1sigDown=cumulative->GetBinCenter(i);
if(bins[i]<0.5)
bandMedian=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(-1))
band1sigUp=cumulative->GetBinCenter(i);
if(bins[i]<RooStats::SignificanceToPValue(-2))
band2sigUp=cumulative->GetBinCenter(i);
}
cout << "-2 sigma band " << band2sigDown << endl;
cout << "-1 sigma band " << band1sigDown << " [Power Constraint)]" << endl;
cout << "median of band " << bandMedian << endl;
cout << "+1 sigma band " << band1sigUp << endl;
cout << "+2 sigma band " << band2sigUp << endl;
// print out the interval on the first Parameter of Interest
cout << "\nobserved 95% upper-limit "<< interval->UpperLimit(*firstPOI) <<endl;
cout << "CLb strict [P(toy>obs|0)] for observed 95% upper-limit "<< CLb <<endl;
cout << "CLb inclusive [P(toy>=obs|0)] for observed 95% upper-limit "<< CLbinclusive <<endl;
delete profile;
delete nll;
}
void ws_constrained_profile3D( const char* wsfile = "rootfiles/ws-data-unblind.root",
const char* new_poi_name = "n_M234_H4_3b",
int npoiPoints = 20,
double poiMinVal = 0.,
double poiMaxVal = 20.,
double constraintWidth = 1.5,
double ymax = 10.,
int verbLevel=0 ) {
gStyle->SetOptStat(0) ;
//--- make output directory.
char command[10000] ;
sprintf( command, "basename %s", wsfile ) ;
TString wsfilenopath = gSystem->GetFromPipe( command ) ;
wsfilenopath.ReplaceAll(".root","") ;
char outputdirstr[1000] ;
sprintf( outputdirstr, "outputfiles/scans-%s", wsfilenopath.Data() ) ;
TString outputdir( outputdirstr ) ;
printf("\n\n Creating output directory: %s\n\n", outputdir.Data() ) ;
sprintf(command, "mkdir -p %s", outputdir.Data() ) ;
gSystem->Exec( command ) ;
//--- Tell RooFit to shut up about anything less important than an ERROR.
RooMsgService::instance().setGlobalKillBelow(RooFit::ERROR) ;
if ( verbLevel > 0 ) { printf("\n\n Verbose level : %d\n\n", verbLevel) ; }
TFile* wstf = new TFile( wsfile ) ;
RooWorkspace* ws = dynamic_cast<RooWorkspace*>( wstf->Get("ws") );
if ( verbLevel > 0 ) { ws->Print() ; }
RooDataSet* rds = (RooDataSet*) ws->obj( "ra2b_observed_rds" ) ;
if ( verbLevel > 0 ) {
printf("\n\n\n ===== RooDataSet ====================\n\n") ;
rds->Print() ;
rds->printMultiline(cout, 1, kTRUE, "") ;
}
ModelConfig* modelConfig = (ModelConfig*) ws->obj( "SbModel" ) ;
RooAbsPdf* likelihood = modelConfig->GetPdf() ;
RooRealVar* rrv_mu_susy_all0lep = ws->var("mu_susy_all0lep") ;
if ( rrv_mu_susy_all0lep == 0x0 ) {
printf("\n\n\n *** can't find mu_susy_all0lep in workspace. Quitting.\n\n\n") ;
return ;
}
//-- do BG only.
rrv_mu_susy_all0lep->setVal(0.) ;
rrv_mu_susy_all0lep->setConstant( kTRUE ) ;
//-- do a prefit.
printf("\n\n\n ====== Pre fit with unmodified nll var.\n\n") ;
RooFitResult* dataFitResultSusyFixed = likelihood->fitTo(*rds, Save(true),Hesse(false),Minos(false),Strategy(1),PrintLevel(verbLevel));
int dataSusyFixedFitCovQual = dataFitResultSusyFixed->covQual() ;
if ( dataSusyFixedFitCovQual < 2 ) { printf("\n\n\n *** Failed fit! Cov qual %d. Quitting.\n\n", dataSusyFixedFitCovQual ) ; return ; }
double dataFitSusyFixedNll = dataFitResultSusyFixed->minNll() ;
if ( verbLevel > 0 ) {
dataFitResultSusyFixed->Print("v") ;
}
printf("\n\n Nll value, from fit result : %.3f\n\n", dataFitSusyFixedNll ) ;
delete dataFitResultSusyFixed ;
//-- Construct the new POI parameter.
RooAbsReal* new_poi_rar(0x0) ;
new_poi_rar = ws->var( new_poi_name ) ;
if ( new_poi_rar == 0x0 ) {
printf("\n\n New POI %s is not a variable. Trying function.\n\n", new_poi_name ) ;
new_poi_rar = ws->function( new_poi_name ) ;
if ( new_poi_rar == 0x0 ) {
printf("\n\n New POI %s is not a function. I quit.\n\n", new_poi_name ) ;
return ;
}
} else {
printf("\n\n New POI %s is a variable with current value %.1f.\n\n", new_poi_name, new_poi_rar->getVal() ) ;
}
if ( npoiPoints <=0 ) {
printf("\n\n Quitting now.\n\n" ) ;
return ;
}
double startPoiVal = new_poi_rar->getVal() ;
//--- The RooNLLVar is NOT equivalent to what minuit uses.
// RooNLLVar* nll = new RooNLLVar("nll","nll", *likelihood, *rds ) ;
// printf("\n\n Nll value, from construction : %.3f\n\n", nll->getVal() ) ;
//--- output of createNLL IS what minuit uses, so use that.
RooAbsReal* nll = likelihood -> createNLL( *rds, Verbose(true) ) ;
RooRealVar* rrv_poiValue = new RooRealVar( "poiValue", "poiValue", 0., -10000., 10000. ) ;
/// rrv_poiValue->setVal( poiMinVal ) ;
/// rrv_poiValue->setConstant(kTRUE) ;
RooRealVar* rrv_constraintWidth = new RooRealVar("constraintWidth","constraintWidth", 0.1, 0.1, 1000. ) ;
rrv_constraintWidth -> setVal( constraintWidth ) ;
rrv_constraintWidth -> setConstant(kTRUE) ;
if ( verbLevel > 0 ) {
printf("\n\n ======= debug likelihood print\n\n") ;
likelihood->Print("v") ;
printf("\n\n ======= debug nll print\n\n") ;
nll->Print("v") ;
}
//----------------------------------------------------------------------------------------------
RooMinuit* rminuit( 0x0 ) ;
RooMinuit* rminuit_uc = new RooMinuit( *nll ) ;
rminuit_uc->setPrintLevel(verbLevel-1) ;
rminuit_uc->setNoWarn() ;
rminuit_uc->migrad() ;
rminuit_uc->hesse() ;
RooFitResult* rfr_uc = rminuit_uc->fit("mr") ;
double floatParInitVal[10000] ;
char floatParName[10000][100] ;
int nFloatParInitVal(0) ;
RooArgList ral_floats = rfr_uc->floatParsFinal() ;
TIterator* floatParIter = ral_floats.createIterator() ;
{
RooRealVar* par ;
while ( (par = (RooRealVar*) floatParIter->Next()) ) {
sprintf( floatParName[nFloatParInitVal], "%s", par->GetName() ) ;
floatParInitVal[nFloatParInitVal] = par->getVal() ;
nFloatParInitVal++ ;
}
}
//-------
printf("\n\n Unbiased best value for new POI %s is : %7.1f\n\n", new_poi_rar->GetName(), new_poi_rar->getVal() ) ;
double best_poi_val = new_poi_rar->getVal() ;
char minuit_formula[10000] ;
sprintf( minuit_formula, "%s+%s*(%s-%s)*(%s-%s)",
nll->GetName(),
rrv_constraintWidth->GetName(),
new_poi_rar->GetName(), rrv_poiValue->GetName(),
new_poi_rar->GetName(), rrv_poiValue->GetName()
) ;
printf("\n\n Creating new minuit variable with formula: %s\n\n", minuit_formula ) ;
RooFormulaVar* new_minuit_var = new RooFormulaVar("new_minuit_var", minuit_formula,
RooArgList( *nll,
*rrv_constraintWidth,
*new_poi_rar, *rrv_poiValue,
*new_poi_rar, *rrv_poiValue
) ) ;
printf("\n\n Current value is %.2f\n\n",
new_minuit_var->getVal() ) ;
rminuit = new RooMinuit( *new_minuit_var ) ;
RooAbsReal* plot_var = nll ;
printf("\n\n Current value is %.2f\n\n",
plot_var->getVal() ) ;
rminuit->setPrintLevel(verbLevel-1) ;
if ( verbLevel <=0 ) { rminuit->setNoWarn() ; }
//----------------------------------------------------------------------------------------------
//-- If POI range is -1 to -1, automatically determine the range using the set value.
if ( poiMinVal < 0. && poiMaxVal < 0. ) {
printf("\n\n Automatic determination of scan range.\n\n") ;
if ( startPoiVal <= 0. ) {
printf("\n\n *** POI starting value zero or negative %g. Quit.\n\n\n", startPoiVal ) ;
return ;
}
poiMinVal = startPoiVal - 3.5 * sqrt(startPoiVal) ;
poiMaxVal = startPoiVal + 6.0 * sqrt(startPoiVal) ;
if ( poiMinVal < 0. ) { poiMinVal = 0. ; }
printf(" Start val = %g. Scan range: %g to %g\n\n", startPoiVal, poiMinVal, poiMaxVal ) ;
}
//----------------------------------------------------------------------------------------------
double poiVals_scanDown[1000] ;
double nllVals_scanDown[1000] ;
//-- Do scan down from best value.
printf("\n\n +++++ Starting scan down from best value.\n\n") ;
double minNllVal(1.e9) ;
for ( int poivi=0; poivi < npoiPoints/2 ; poivi++ ) {
////double poiValue = poiMinVal + poivi*(poiMaxVal-poiMinVal)/(1.*(npoiPoints-1)) ;
double poiValue = best_poi_val - poivi*(best_poi_val-poiMinVal)/(1.*(npoiPoints/2-1)) ;
rrv_poiValue -> setVal( poiValue ) ;
rrv_poiValue -> setConstant( kTRUE ) ;
//+++++++++++++++++++++++++++++++++++
rminuit->migrad() ;
rminuit->hesse() ;
RooFitResult* rfr = rminuit->save() ;
//+++++++++++++++++++++++++++++++++++
if ( verbLevel > 0 ) { rfr->Print("v") ; }
float fit_minuit_var_val = rfr->minNll() ;
printf(" %02d : poi constraint = %.2f : allvars : MinuitVar, createNLL, PV, POI : %.5f %.5f %.5f %.5f\n",
poivi, rrv_poiValue->getVal(), fit_minuit_var_val, nll->getVal(), plot_var->getVal(), new_poi_rar->getVal() ) ;
cout << flush ;
poiVals_scanDown[poivi] = new_poi_rar->getVal() ;
nllVals_scanDown[poivi] = plot_var->getVal() ;
if ( nllVals_scanDown[poivi] < minNllVal ) { minNllVal = nllVals_scanDown[poivi] ; }
delete rfr ;
} // poivi
printf("\n\n +++++ Resetting floats to best fit values.\n\n") ;
for ( int pi=0; pi<nFloatParInitVal; pi++ ) {
RooRealVar* par = ws->var( floatParName[pi] ) ;
par->setVal( floatParInitVal[pi] ) ;
} // pi.
printf("\n\n +++++ Starting scan up from best value.\n\n") ;
//-- Now do scan up.
double poiVals_scanUp[1000] ;
double nllVals_scanUp[1000] ;
for ( int poivi=0; poivi < npoiPoints/2 ; poivi++ ) {
double poiValue = best_poi_val + poivi*(poiMaxVal-best_poi_val)/(1.*(npoiPoints/2-1)) ;
rrv_poiValue -> setVal( poiValue ) ;
rrv_poiValue -> setConstant( kTRUE ) ;
//+++++++++++++++++++++++++++++++++++
rminuit->migrad() ;
rminuit->hesse() ;
RooFitResult* rfr = rminuit->save() ;
//+++++++++++++++++++++++++++++++++++
if ( verbLevel > 0 ) { rfr->Print("v") ; }
float fit_minuit_var_val = rfr->minNll() ;
printf(" %02d : poi constraint = %.2f : allvars : MinuitVar, createNLL, PV, POI : %.5f %.5f %.5f %.5f\n",
poivi, rrv_poiValue->getVal(), fit_minuit_var_val, nll->getVal(), plot_var->getVal(), new_poi_rar->getVal() ) ;
cout << flush ;
poiVals_scanUp[poivi] = new_poi_rar->getVal() ;
nllVals_scanUp[poivi] = plot_var->getVal() ;
if ( nllVals_scanUp[poivi] < minNllVal ) { minNllVal = nllVals_scanUp[poivi] ; }
delete rfr ;
} // poivi
double poiVals[1000] ;
double nllVals[1000] ;
int pointCount(0) ;
for ( int pi=0; pi<npoiPoints/2; pi++ ) {
poiVals[pi] = poiVals_scanDown[(npoiPoints/2-1)-pi] ;
nllVals[pi] = nllVals_scanDown[(npoiPoints/2-1)-pi] ;
pointCount++ ;
}
for ( int pi=1; pi<npoiPoints/2; pi++ ) {
poiVals[pointCount] = poiVals_scanUp[pi] ;
nllVals[pointCount] = nllVals_scanUp[pi] ;
pointCount++ ;
}
npoiPoints = pointCount ;
printf("\n\n --- TGraph arrays:\n") ;
for ( int i=0; i<npoiPoints; i++ ) {
printf(" %2d : poi = %6.1f, nll = %g\n", i, poiVals[i], nllVals[i] ) ;
}
printf("\n\n") ;
double nllDiffVals[1000] ;
double poiAtMinlnL(-1.) ;
double poiAtMinusDelta2(-1.) ;
double poiAtPlusDelta2(-1.) ;
for ( int poivi=0; poivi < npoiPoints ; poivi++ ) {
nllDiffVals[poivi] = 2.*(nllVals[poivi] - minNllVal) ;
double poiValue = poiMinVal + poivi*(poiMaxVal-poiMinVal)/(1.*npoiPoints) ;
if ( nllDiffVals[poivi] < 0.01 ) { poiAtMinlnL = poiValue ; }
if ( poiAtMinusDelta2 < 0. && nllDiffVals[poivi] < 2.5 ) { poiAtMinusDelta2 = poiValue ; }
if ( poiAtMinlnL > 0. && poiAtPlusDelta2 < 0. && nllDiffVals[poivi] > 2.0 ) { poiAtPlusDelta2 = poiValue ; }
} // poivi
printf("\n\n Estimates for poi at delta ln L = -2, 0, +2: %g , %g , %g\n\n", poiAtMinusDelta2, poiAtMinlnL, poiAtPlusDelta2 ) ;
//--- Main canvas
TCanvas* cscan = (TCanvas*) gDirectory->FindObject("cscan") ;
if ( cscan == 0x0 ) {
printf("\n Creating canvas.\n\n") ;
cscan = new TCanvas("cscan","Delta nll") ;
}
char gname[1000] ;
TGraph* graph = new TGraph( npoiPoints, poiVals, nllDiffVals ) ;
sprintf( gname, "scan_%s", new_poi_name ) ;
graph->SetName( gname ) ;
double poiBest(-1.) ;
double poiMinus1stdv(-1.) ;
double poiPlus1stdv(-1.) ;
double poiMinus2stdv(-1.) ;
double poiPlus2stdv(-1.) ;
double twoDeltalnLMin(1e9) ;
int nscan(1000) ;
for ( int xi=0; xi<nscan; xi++ ) {
double x = poiVals[0] + xi*(poiVals[npoiPoints-1]-poiVals[0])/(nscan-1) ;
double twoDeltalnL = graph -> Eval( x, 0, "S" ) ;
if ( poiMinus1stdv < 0. && twoDeltalnL < 1.0 ) { poiMinus1stdv = x ; printf(" set m1 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( poiMinus2stdv < 0. && twoDeltalnL < 4.0 ) { poiMinus2stdv = x ; printf(" set m2 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( twoDeltalnL < twoDeltalnLMin ) { poiBest = x ; twoDeltalnLMin = twoDeltalnL ; }
if ( twoDeltalnLMin < 0.3 && poiPlus1stdv < 0. && twoDeltalnL > 1.0 ) { poiPlus1stdv = x ; printf(" set p1 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( twoDeltalnLMin < 0.3 && poiPlus2stdv < 0. && twoDeltalnL > 4.0 ) { poiPlus2stdv = x ; printf(" set p2 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( xi%100 == 0 ) { printf( " %4d : poi=%6.2f, 2DeltalnL = %6.2f\n", xi, x, twoDeltalnL ) ; }
}
printf("\n\n POI estimate : %g +%g -%g [%g,%g], two sigma errors: +%g -%g [%g,%g]\n\n",
poiBest,
(poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv), poiMinus1stdv, poiPlus1stdv,
(poiPlus2stdv-poiBest), (poiBest-poiMinus2stdv), poiMinus2stdv, poiPlus2stdv
) ;
printf(" %s val,pm1sig,pm2sig: %7.2f %7.2f %7.2f %7.2f %7.2f\n",
new_poi_name, poiBest, (poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv), (poiPlus2stdv-poiBest), (poiBest-poiMinus2stdv) ) ;
char htitle[1000] ;
sprintf(htitle, "%s profile likelihood scan: -2ln(L/Lm)", new_poi_name ) ;
TH1F* hscan = new TH1F("hscan", htitle, 10, poiMinVal, poiMaxVal ) ;
hscan->SetMinimum(0.) ;
hscan->SetMaximum(ymax) ;
hscan->DrawCopy() ;
graph->SetLineColor(4) ;
graph->SetLineWidth(3) ;
graph->Draw("CP") ;
gPad->SetGridx(1) ;
gPad->SetGridy(1) ;
cscan->Update() ;
TLine* line = new TLine() ;
line->SetLineColor(2) ;
line->DrawLine(poiMinVal, 1., poiPlus1stdv, 1.) ;
line->DrawLine(poiMinus1stdv,0., poiMinus1stdv, 1.) ;
line->DrawLine(poiPlus1stdv ,0., poiPlus1stdv , 1.) ;
TText* text = new TText() ;
text->SetTextSize(0.04) ;
char tstring[1000] ;
sprintf( tstring, "%s = %.1f +%.1f -%.1f", new_poi_name, poiBest, (poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv) ) ;
text -> DrawTextNDC( 0.15, 0.85, tstring ) ;
sprintf( tstring, "68%% interval [%.1f, %.1f]", poiMinus1stdv, poiPlus1stdv ) ;
text -> DrawTextNDC( 0.15, 0.78, tstring ) ;
char hname[1000] ;
sprintf( hname, "hscanout_%s", new_poi_name ) ;
TH1F* hsout = new TH1F( hname,"scan results",4,0.,4.) ;
double obsVal(-1.) ;
hsout->SetBinContent(1, obsVal ) ;
hsout->SetBinContent(2, poiPlus1stdv ) ;
hsout->SetBinContent(3, poiBest ) ;
hsout->SetBinContent(4, poiMinus1stdv ) ;
TAxis* xaxis = hsout->GetXaxis() ;
xaxis->SetBinLabel(1,"Observed val.") ;
xaxis->SetBinLabel(2,"Model+1sd") ;
xaxis->SetBinLabel(3,"Model") ;
xaxis->SetBinLabel(4,"Model-1sd") ;
char outrootfile[10000] ;
sprintf( outrootfile, "%s/scan-ff-%s.root", outputdir.Data(), new_poi_name ) ;
char outpdffile[10000] ;
sprintf( outpdffile, "%s/scan-ff-%s.pdf", outputdir.Data(), new_poi_name ) ;
cscan->Update() ; cscan->Draw() ;
printf("\n Saving %s\n", outpdffile ) ;
cscan->SaveAs( outpdffile ) ;
//--- save in root file
printf("\n Saving %s\n", outrootfile ) ;
TFile fout(outrootfile,"recreate") ;
graph->Write() ;
hsout->Write() ;
fout.Close() ;
delete ws ;
wstf->Close() ;
}
void StandardFeldmanCousinsDemo(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
// -------------------------------------------------------
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
// -------------------------------------------------------
// Tutorial starts here
// -------------------------------------------------------
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
// -------------------------------------------------------
// create and use the FeldmanCousins tool
// to find and plot the 95% confidence interval
// on the parameter of interest as specified
// in the model config
FeldmanCousins fc(*data,*mc);
fc.SetConfidenceLevel(0.95); // 95% interval
//fc.AdditionalNToysFactor(0.1); // to speed up the result
fc.UseAdaptiveSampling(true); // speed it up a bit
fc.SetNBins(10); // set how many points per parameter of interest to scan
fc.CreateConfBelt(true); // save the information in the belt for plotting
// Since this tool needs to throw toy MC the PDF needs to be
// extended or the tool needs to know how many entries in a dataset
// per pseudo experiment.
// In the 'number counting form' where the entries in the dataset
// are counts, and not values of discriminating variables, the
// datasets typically only have one entry and the PDF is not
// extended.
if(!mc->GetPdf()->canBeExtended()){
if(data->numEntries()==1)
fc.FluctuateNumDataEntries(false);
else
cout <<"Not sure what to do about this model" <<endl;
}
// We can use PROOF to speed things along in parallel
// ProofConfig pc(*w, 1, "workers=4", kFALSE);
// ToyMCSampler* toymcsampler = (ToyMCSampler*) fc.GetTestStatSampler();
// toymcsampler->SetProofConfig(&pc); // enable proof
// Now get the interval
PointSetInterval* interval = fc.GetInterval();
ConfidenceBelt* belt = fc.GetConfidenceBelt();
// print out the iterval on the first Parameter of Interest
RooRealVar* firstPOI = (RooRealVar*) mc->GetParametersOfInterest()->first();
cout << "\n95% interval on " <<firstPOI->GetName()<<" is : ["<<
interval->LowerLimit(*firstPOI) << ", "<<
interval->UpperLimit(*firstPOI) <<"] "<<endl;
// ---------------------------------------------
// No nice plots yet, so plot the belt by hand
// Ask the calculator which points were scanned
RooDataSet* parameterScan = (RooDataSet*) fc.GetPointsToScan();
RooArgSet* tmpPoint;
// make a histogram of parameter vs. threshold
TH1F* histOfThresholds = new TH1F("histOfThresholds","",
parameterScan->numEntries(),
firstPOI->getMin(),
firstPOI->getMax());
// loop through the points that were tested and ask confidence belt
// what the upper/lower thresholds were.
// For FeldmanCousins, the lower cut off is always 0
for(Int_t i=0; i<parameterScan->numEntries(); ++i){
tmpPoint = (RooArgSet*) parameterScan->get(i)->clone("temp");
double arMax = belt->GetAcceptanceRegionMax(*tmpPoint);
double arMin = belt->GetAcceptanceRegionMax(*tmpPoint);
double poiVal = tmpPoint->getRealValue(firstPOI->GetName()) ;
histOfThresholds->Fill(poiVal,arMax);
}
histOfThresholds->SetMinimum(0);
histOfThresholds->Draw();
}
void StandardHistFactoryPlotsWithCategories(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData"){
double nSigmaToVary=5.;
double muVal=0;
bool doFit=false;
// -------------------------------------------------------
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
// -------------------------------------------------------
// Tutorial starts here
// -------------------------------------------------------
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
// -------------------------------------------------------
// now use the profile inspector
RooRealVar* obs = (RooRealVar*)mc->GetObservables()->first();
TList* list = new TList();
RooRealVar * firstPOI = dynamic_cast<RooRealVar*>(mc->GetParametersOfInterest()->first());
firstPOI->setVal(muVal);
// firstPOI->setConstant();
if(doFit){
mc->GetPdf()->fitTo(*data);
}
// -------------------------------------------------------
mc->GetNuisanceParameters()->Print("v");
int nPlotsMax = 1000;
cout <<" check expectedData by category"<<endl;
RooDataSet* simData=NULL;
RooSimultaneous* simPdf = NULL;
if(strcmp(mc->GetPdf()->ClassName(),"RooSimultaneous")==0){
cout <<"Is a simultaneous PDF"<<endl;
simPdf = (RooSimultaneous *)(mc->GetPdf());
} else {
cout <<"Is not a simultaneous PDF"<<endl;
}
if(doFit) {
RooCategory* channelCat = (RooCategory*) (&simPdf->indexCat());
TIterator* iter = channelCat->typeIterator() ;
RooCatType* tt = NULL;
tt=(RooCatType*) iter->Next();
RooAbsPdf* pdftmp = ((RooSimultaneous*)mc->GetPdf())->getPdf(tt->GetName()) ;
RooArgSet* obstmp = pdftmp->getObservables(*mc->GetObservables()) ;
obs = ((RooRealVar*)obstmp->first());
RooPlot* frame = obs->frame();
cout <<Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())<<endl;
cout << tt->GetName() << " " << channelCat->getLabel() <<endl;
data->plotOn(frame,MarkerSize(1),Cut(Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())),DataError(RooAbsData::None));
Double_t normCount = data->sumEntries(Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())) ;
pdftmp->plotOn(frame,LineWidth(2.),Normalization(normCount,RooAbsReal::NumEvent)) ;
frame->Draw();
cout <<"expected events = " << mc->GetPdf()->expectedEvents(*data->get()) <<endl;
return;
}
int nPlots=0;
if(!simPdf){
TIterator* it = mc->GetNuisanceParameters()->createIterator();
RooRealVar* var = NULL;
while( (var = (RooRealVar*) it->Next()) != NULL){
RooPlot* frame = obs->frame();
frame->SetYTitle(var->GetName());
data->plotOn(frame,MarkerSize(1));
var->setVal(0);
mc->GetPdf()->plotOn(frame,LineWidth(1.));
var->setVal(1);
mc->GetPdf()->plotOn(frame,LineColor(kRed),LineStyle(kDashed),LineWidth(1));
var->setVal(-1);
mc->GetPdf()->plotOn(frame,LineColor(kGreen),LineStyle(kDashed),LineWidth(1));
list->Add(frame);
var->setVal(0);
}
} else {
RooCategory* channelCat = (RooCategory*) (&simPdf->indexCat());
// TIterator* iter = simPdf->indexCat().typeIterator() ;
TIterator* iter = channelCat->typeIterator() ;
RooCatType* tt = NULL;
while(nPlots<nPlotsMax && (tt=(RooCatType*) iter->Next())) {
cout << "on type " << tt->GetName() << " " << endl;
// Get pdf associated with state from simpdf
RooAbsPdf* pdftmp = simPdf->getPdf(tt->GetName()) ;
// Generate observables defined by the pdf associated with this state
RooArgSet* obstmp = pdftmp->getObservables(*mc->GetObservables()) ;
// obstmp->Print();
obs = ((RooRealVar*)obstmp->first());
TIterator* it = mc->GetNuisanceParameters()->createIterator();
RooRealVar* var = NULL;
while(nPlots<nPlotsMax && (var = (RooRealVar*) it->Next())){
TCanvas* c2 = new TCanvas("c2");
RooPlot* frame = obs->frame();
frame->SetName(Form("frame%d",nPlots));
frame->SetYTitle(var->GetName());
cout <<Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())<<endl;
cout << tt->GetName() << " " << channelCat->getLabel() <<endl;
data->plotOn(frame,MarkerSize(1),Cut(Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())),DataError(RooAbsData::None));
Double_t normCount = data->sumEntries(Form("%s==%s::%s",channelCat->GetName(),channelCat->GetName(),tt->GetName())) ;
if(strcmp(var->GetName(),"Lumi")==0){
cout <<"working on lumi"<<endl;
var->setVal(w->var("nominalLumi")->getVal());
var->Print();
} else{
var->setVal(0);
}
// w->allVars().Print("v");
// mc->GetNuisanceParameters()->Print("v");
// pdftmp->plotOn(frame,LineWidth(2.));
// mc->GetPdf()->plotOn(frame,LineWidth(2.),Slice(*channelCat,tt->GetName()),ProjWData(*data));
//pdftmp->plotOn(frame,LineWidth(2.),Slice(*channelCat,tt->GetName()),ProjWData(*data));
normCount = pdftmp->expectedEvents(*obs);
pdftmp->plotOn(frame,LineWidth(2.),Normalization(normCount,RooAbsReal::NumEvent)) ;
if(strcmp(var->GetName(),"Lumi")==0){
cout <<"working on lumi"<<endl;
var->setVal(w->var("nominalLumi")->getVal()+0.05);
var->Print();
} else{
var->setVal(nSigmaToVary);
}
// pdftmp->plotOn(frame,LineColor(kRed),LineStyle(kDashed),LineWidth(2));
// mc->GetPdf()->plotOn(frame,LineColor(kRed),LineStyle(kDashed),LineWidth(2.),Slice(*channelCat,tt->GetName()),ProjWData(*data));
//pdftmp->plotOn(frame,LineColor(kRed),LineStyle(kDashed),LineWidth(2.),Slice(*channelCat,tt->GetName()),ProjWData(*data));
normCount = pdftmp->expectedEvents(*obs);
pdftmp->plotOn(frame,LineWidth(2.),LineColor(kRed),LineStyle(kDashed),Normalization(normCount,RooAbsReal::NumEvent)) ;
if(strcmp(var->GetName(),"Lumi")==0){
cout <<"working on lumi"<<endl;
var->setVal(w->var("nominalLumi")->getVal()-0.05);
var->Print();
} else{
var->setVal(-nSigmaToVary);
}
// pdftmp->plotOn(frame,LineColor(kGreen),LineStyle(kDashed),LineWidth(2));
// mc->GetPdf()->plotOn(frame,LineColor(kGreen),LineStyle(kDashed),LineWidth(2),Slice(*channelCat,tt->GetName()),ProjWData(*data));
//pdftmp->plotOn(frame,LineColor(kGreen),LineStyle(kDashed),LineWidth(2),Slice(*channelCat,tt->GetName()),ProjWData(*data));
normCount = pdftmp->expectedEvents(*obs);
pdftmp->plotOn(frame,LineWidth(2.),LineColor(kGreen),LineStyle(kDashed),Normalization(normCount,RooAbsReal::NumEvent)) ;
// set them back to normal
if(strcmp(var->GetName(),"Lumi")==0){
cout <<"working on lumi"<<endl;
var->setVal(w->var("nominalLumi")->getVal());
var->Print();
} else{
var->setVal(0);
}
list->Add(frame);
// quit making plots
++nPlots;
frame->Draw();
c2->SaveAs(Form("%s_%s_%s.pdf",tt->GetName(),obs->GetName(),var->GetName()));
delete c2;
}
}
}
// -------------------------------------------------------
// now make plots
TCanvas* c1 = new TCanvas("c1","ProfileInspectorDemo",800,200);
if(list->GetSize()>4){
double n = list->GetSize();
int nx = (int)sqrt(n) ;
int ny = TMath::CeilNint(n/nx);
nx = TMath::CeilNint( sqrt(n) );
c1->Divide(ny,nx);
} else
c1->Divide(list->GetSize());
for(int i=0; i<list->GetSize(); ++i){
c1->cd(i+1);
list->At(i)->Draw();
}
}
//put very small data entries in a binned dataset to avoid unphysical pdfs, specifically for H->ZZ->4l
RooDataSet* makeData(RooDataSet* orig, RooSimultaneous* simPdf, const RooArgSet* observables, RooRealVar* firstPOI, double mass, double& mu_min)
{
double max_soverb = 0;
mu_min = -10e9;
map<string, RooDataSet*> data_map;
firstPOI->setVal(0);
RooCategory* cat = (RooCategory*)&simPdf->indexCat();
TList* datalist = orig->split(*(RooAbsCategory*)cat, true);
TIterator* dataItr = datalist->MakeIterator();
RooAbsData* ds;
RooRealVar* weightVar = new RooRealVar("weightVar","weightVar",1);
while ((ds = (RooAbsData*)dataItr->Next()))
{
string typeName(ds->GetName());
cat->setLabel(typeName.c_str());
RooAbsPdf* pdf = simPdf->getPdf(typeName.c_str());
cout << "pdf: " << pdf << endl;
RooArgSet* obs = pdf->getObservables(observables);
cout << "obs: " << obs << endl;
RooArgSet obsAndWeight(*obs, *weightVar);
obsAndWeight.add(*cat);
stringstream datasetName;
datasetName << "newData_" << typeName;
RooDataSet* thisData = new RooDataSet(datasetName.str().c_str(),datasetName.str().c_str(), obsAndWeight, WeightVar(*weightVar));
RooRealVar* firstObs = (RooRealVar*)obs->first();
//int ibin = 0;
int nrEntries = ds->numEntries();
for (int ib=0;ib<nrEntries;ib++)
{
const RooArgSet* event = ds->get(ib);
const RooRealVar* thisObs = (RooRealVar*)event->find(firstObs->GetName());
firstObs->setVal(thisObs->getVal());
firstPOI->setVal(0);
double b = pdf->expectedEvents(*firstObs)*pdf->getVal(obs);
firstPOI->setVal(1);
double s = pdf->expectedEvents(*firstObs)*pdf->getVal(obs) - b;
if (s > 0)
{
mu_min = max(mu_min, -b/s);
double soverb = s/b;
if (soverb > max_soverb)
{
max_soverb = soverb;
cout << "Found new max s/b: " << soverb << " in pdf " << pdf->GetName() << " at m = " << thisObs->getVal() << endl;
}
}
if (b == 0 && s != 0)
{
cout << "Expecting non-zero signal and zero bg at m=" << firstObs->getVal() << " in pdf " << pdf->GetName() << endl;
}
if (s+b <= 0)
{
cout << "expecting zero" << endl;
continue;
}
double weight = ds->weight();
if ((typeName.find("ATLAS_H_4mu") != string::npos ||
typeName.find("ATLAS_H_4e") != string::npos ||
typeName.find("ATLAS_H_2mu2e") != string::npos ||
typeName.find("ATLAS_H_2e2mu") != string::npos) && fabs(firstObs->getVal() - mass) < 10 && weight == 0)
{
cout << "adding event: " << firstObs->getVal() << endl;
thisData->add(*event, pow(10., -9.));
}
else
{
//weight = max(pow(10.0, -9), weight);
thisData->add(*event, weight);
}
}
data_map[string(ds->GetName())] = (RooDataSet*)thisData;
}
RooDataSet* newData = new RooDataSet("newData","newData",RooArgSet(*observables, *weightVar),
Index(*cat), Import(data_map), WeightVar(*weightVar));
orig->Print();
newData->Print();
//newData->tree()->Scan("*");
return newData;
}
void constrained_scan( const char* wsfile = "outputfiles/ws.root",
const char* new_poi_name="mu_bg_4b_msig_met1",
double constraintWidth=1.5,
int npoiPoints = 20,
double poiMinVal = 0.,
double poiMaxVal = 10.0,
double ymax = 9.,
int verbLevel=1 ) {
TString outputdir("outputfiles") ;
gStyle->SetOptStat(0) ;
TFile* wstf = new TFile( wsfile ) ;
RooWorkspace* ws = dynamic_cast<RooWorkspace*>( wstf->Get("ws") );
ws->Print() ;
RooDataSet* rds = (RooDataSet*) ws->obj( "hbb_observed_rds" ) ;
cout << "\n\n\n ===== RooDataSet ====================\n\n" << endl ;
rds->Print() ;
rds->printMultiline(cout, 1, kTRUE, "") ;
RooRealVar* rv_sig_strength = ws->var("sig_strength") ;
if ( rv_sig_strength == 0x0 ) { printf("\n\n *** can't find sig_strength in workspace.\n\n" ) ; return ; }
RooAbsPdf* likelihood = ws->pdf("likelihood") ;
if ( likelihood == 0x0 ) { printf("\n\n *** can't find likelihood in workspace.\n\n" ) ; return ; }
printf("\n\n Likelihood:\n") ;
likelihood -> Print() ;
/////rv_sig_strength -> setConstant( kFALSE ) ;
rv_sig_strength -> setVal(0.) ;
rv_sig_strength -> setConstant( kTRUE ) ;
likelihood->fitTo( *rds, Save(false), PrintLevel(0), Hesse(true), Strategy(1) ) ;
//RooFitResult* fitResult = likelihood->fitTo( *rds, Save(true), PrintLevel(0), Hesse(true), Strategy(1) ) ;
//double minNllSusyFloat = fitResult->minNll() ;
//double susy_ss_atMinNll = rv_sig_strength -> getVal() ;
RooMsgService::instance().getStream(1).removeTopic(Minimization) ;
RooMsgService::instance().getStream(1).removeTopic(Fitting) ;
//-- Construct the new POI parameter.
RooAbsReal* new_poi_rar(0x0) ;
new_poi_rar = ws->var( new_poi_name ) ;
if ( new_poi_rar == 0x0 ) {
printf("\n\n New POI %s is not a variable. Trying function.\n\n", new_poi_name ) ;
new_poi_rar = ws->function( new_poi_name ) ;
if ( new_poi_rar == 0x0 ) {
printf("\n\n New POI %s is not a function. I quit.\n\n", new_poi_name ) ;
return ;
} else {
printf("\n Found it.\n\n") ;
}
} else {
printf("\n\n New POI %s is a variable with current value %.1f.\n\n", new_poi_name, new_poi_rar->getVal() ) ;
}
double startPoiVal = new_poi_rar->getVal() ;
RooAbsReal* nll = likelihood -> createNLL( *rds, Verbose(true) ) ;
RooRealVar* rrv_poiValue = new RooRealVar( "poiValue", "poiValue", 0., -10000., 10000. ) ;
RooRealVar* rrv_constraintWidth = new RooRealVar("constraintWidth","constraintWidth", 0.1, 0.1, 1000. ) ;
rrv_constraintWidth -> setVal( constraintWidth ) ;
rrv_constraintWidth -> setConstant(kTRUE) ;
RooMinuit* rminuit( 0x0 ) ;
RooMinuit* rminuit_uc = new RooMinuit( *nll ) ;
rminuit_uc->setPrintLevel(verbLevel-1) ;
rminuit_uc->setNoWarn() ;
rminuit_uc->migrad() ;
rminuit_uc->hesse() ;
RooFitResult* rfr_uc = rminuit_uc->fit("mr") ;
double floatParInitVal[10000] ;
char floatParName[10000][100] ;
int nFloatParInitVal(0) ;
RooArgList ral_floats = rfr_uc->floatParsFinal() ;
TIterator* floatParIter = ral_floats.createIterator() ;
{
RooRealVar* par ;
while ( (par = (RooRealVar*) floatParIter->Next()) ) {
sprintf( floatParName[nFloatParInitVal], "%s", par->GetName() ) ;
floatParInitVal[nFloatParInitVal] = par->getVal() ;
nFloatParInitVal++ ;
}
}
printf("\n\n Unbiased best value for new POI %s is : %7.1f\n\n", new_poi_rar->GetName(), new_poi_rar->getVal() ) ;
double best_poi_val = new_poi_rar->getVal() ;
char minuit_formula[10000] ;
sprintf( minuit_formula, "%s+%s*(%s-%s)*(%s-%s)",
nll->GetName(),
rrv_constraintWidth->GetName(),
new_poi_rar->GetName(), rrv_poiValue->GetName(),
new_poi_rar->GetName(), rrv_poiValue->GetName()
) ;
printf("\n\n Creating new minuit variable with formula: %s\n\n", minuit_formula ) ;
RooFormulaVar* new_minuit_var = new RooFormulaVar("new_minuit_var", minuit_formula,
RooArgList( *nll,
*rrv_constraintWidth,
*new_poi_rar, *rrv_poiValue,
*new_poi_rar, *rrv_poiValue
) ) ;
printf("\n\n Current value is %.2f\n\n",
new_minuit_var->getVal() ) ;
rminuit = new RooMinuit( *new_minuit_var ) ;
RooAbsReal* plot_var = nll ;
printf("\n\n Current value is %.2f\n\n",
plot_var->getVal() ) ;
rminuit->setPrintLevel(verbLevel-1) ;
if ( verbLevel <=0 ) { rminuit->setNoWarn() ; }
if ( poiMinVal < 0. && poiMaxVal < 0. ) {
printf("\n\n Automatic determination of scan range.\n\n") ;
if ( startPoiVal <= 0. ) {
printf("\n\n *** POI starting value zero or negative %g. Quit.\n\n\n", startPoiVal ) ;
return ;
}
poiMinVal = startPoiVal - 3.5 * sqrt(startPoiVal) ;
poiMaxVal = startPoiVal + 6.0 * sqrt(startPoiVal) ;
if ( poiMinVal < 0. ) { poiMinVal = 0. ; }
printf(" Start val = %g. Scan range: %g to %g\n\n", startPoiVal, poiMinVal, poiMaxVal ) ;
}
//----------------------------------------------------------------------------------------------
double poiVals_scanDown[1000] ;
double nllVals_scanDown[1000] ;
//-- Do scan down from best value.
printf("\n\n +++++ Starting scan down from best value.\n\n") ;
double minNllVal(1.e9) ;
for ( int poivi=0; poivi < npoiPoints/2 ; poivi++ ) {
////double poiValue = poiMinVal + poivi*(poiMaxVal-poiMinVal)/(1.*(npoiPoints-1)) ;
double poiValue = best_poi_val - poivi*(best_poi_val-poiMinVal)/(1.*(npoiPoints/2-1)) ;
rrv_poiValue -> setVal( poiValue ) ;
rrv_poiValue -> setConstant( kTRUE ) ;
//+++++++++++++++++++++++++++++++++++
rminuit->migrad() ;
rminuit->hesse() ;
RooFitResult* rfr = rminuit->save() ;
//+++++++++++++++++++++++++++++++++++
if ( verbLevel > 0 ) { rfr->Print("v") ; }
float fit_minuit_var_val = rfr->minNll() ;
printf(" %02d : poi constraint = %.2f : allvars : MinuitVar, createNLL, PV, POI : %.5f %.5f %.5f %.5f\n",
poivi, rrv_poiValue->getVal(), fit_minuit_var_val, nll->getVal(), plot_var->getVal(), new_poi_rar->getVal() ) ;
cout << flush ;
poiVals_scanDown[poivi] = new_poi_rar->getVal() ;
nllVals_scanDown[poivi] = plot_var->getVal() ;
if ( nllVals_scanDown[poivi] < minNllVal ) { minNllVal = nllVals_scanDown[poivi] ; }
delete rfr ;
} // poivi
printf("\n\n +++++ Resetting floats to best fit values.\n\n") ;
for ( int pi=0; pi<nFloatParInitVal; pi++ ) {
RooRealVar* par = ws->var( floatParName[pi] ) ;
par->setVal( floatParInitVal[pi] ) ;
} // pi.
printf("\n\n +++++ Starting scan up from best value.\n\n") ;
//-- Now do scan up.
double poiVals_scanUp[1000] ;
double nllVals_scanUp[1000] ;
for ( int poivi=0; poivi < npoiPoints/2 ; poivi++ ) {
double poiValue = best_poi_val + poivi*(poiMaxVal-best_poi_val)/(1.*(npoiPoints/2-1)) ;
rrv_poiValue -> setVal( poiValue ) ;
rrv_poiValue -> setConstant( kTRUE ) ;
//+++++++++++++++++++++++++++++++++++
rminuit->migrad() ;
rminuit->hesse() ;
RooFitResult* rfr = rminuit->save() ;
//+++++++++++++++++++++++++++++++++++
if ( verbLevel > 0 ) { rfr->Print("v") ; }
float fit_minuit_var_val = rfr->minNll() ;
printf(" %02d : poi constraint = %.2f : allvars : MinuitVar, createNLL, PV, POI : %.5f %.5f %.5f %.5f\n",
poivi, rrv_poiValue->getVal(), fit_minuit_var_val, nll->getVal(), plot_var->getVal(), new_poi_rar->getVal() ) ;
cout << flush ;
poiVals_scanUp[poivi] = new_poi_rar->getVal() ;
nllVals_scanUp[poivi] = plot_var->getVal() ;
if ( nllVals_scanUp[poivi] < minNllVal ) { minNllVal = nllVals_scanUp[poivi] ; }
delete rfr ;
} // poivi
double poiVals[1000] ;
double nllVals[1000] ;
int pointCount(0) ;
for ( int pi=0; pi<npoiPoints/2; pi++ ) {
poiVals[pi] = poiVals_scanDown[(npoiPoints/2-1)-pi] ;
nllVals[pi] = nllVals_scanDown[(npoiPoints/2-1)-pi] ;
pointCount++ ;
}
for ( int pi=1; pi<npoiPoints/2; pi++ ) {
poiVals[pointCount] = poiVals_scanUp[pi] ;
nllVals[pointCount] = nllVals_scanUp[pi] ;
pointCount++ ;
}
npoiPoints = pointCount ;
printf("\n\n --- TGraph arrays:\n") ;
for ( int i=0; i<npoiPoints; i++ ) {
printf(" %2d : poi = %6.1f, nll = %g\n", i, poiVals[i], nllVals[i] ) ;
}
printf("\n\n") ;
double nllDiffVals[1000] ;
double poiAtMinlnL(-1.) ;
double poiAtMinusDelta2(-1.) ;
double poiAtPlusDelta2(-1.) ;
for ( int poivi=0; poivi < npoiPoints ; poivi++ ) {
nllDiffVals[poivi] = 2.*(nllVals[poivi] - minNllVal) ;
double poiValue = poiMinVal + poivi*(poiMaxVal-poiMinVal)/(1.*npoiPoints) ;
if ( nllDiffVals[poivi] < 0.01 ) { poiAtMinlnL = poiValue ; }
if ( poiAtMinusDelta2 < 0. && nllDiffVals[poivi] < 2.5 ) { poiAtMinusDelta2 = poiValue ; }
if ( poiAtMinlnL > 0. && poiAtPlusDelta2 < 0. && nllDiffVals[poivi] > 2.0 ) { poiAtPlusDelta2 = poiValue ; }
} // poivi
printf("\n\n Estimates for poi at delta ln L = -2, 0, +2: %g , %g , %g\n\n", poiAtMinusDelta2, poiAtMinlnL, poiAtPlusDelta2 ) ;
//--- Main canvas
TCanvas* cscan = (TCanvas*) gDirectory->FindObject("cscan") ;
if ( cscan == 0x0 ) {
printf("\n Creating canvas.\n\n") ;
cscan = new TCanvas("cscan","Delta nll") ;
}
char gname[1000] ;
TGraph* graph = new TGraph( npoiPoints, poiVals, nllDiffVals ) ;
sprintf( gname, "scan_%s", new_poi_name ) ;
graph->SetName( gname ) ;
double poiBest(-1.) ;
double poiMinus1stdv(-1.) ;
double poiPlus1stdv(-1.) ;
double poiMinus2stdv(-1.) ;
double poiPlus2stdv(-1.) ;
double twoDeltalnLMin(1e9) ;
int nscan(1000) ;
for ( int xi=0; xi<nscan; xi++ ) {
double x = poiVals[0] + xi*(poiVals[npoiPoints-1]-poiVals[0])/(nscan-1) ;
double twoDeltalnL = graph -> Eval( x, 0, "S" ) ;
if ( poiMinus1stdv < 0. && twoDeltalnL < 1.0 ) { poiMinus1stdv = x ; printf(" set m1 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( poiMinus2stdv < 0. && twoDeltalnL < 4.0 ) { poiMinus2stdv = x ; printf(" set m2 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( twoDeltalnL < twoDeltalnLMin ) { poiBest = x ; twoDeltalnLMin = twoDeltalnL ; }
if ( twoDeltalnLMin < 0.3 && poiPlus1stdv < 0. && twoDeltalnL > 1.0 ) { poiPlus1stdv = x ; printf(" set p1 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( twoDeltalnLMin < 0.3 && poiPlus2stdv < 0. && twoDeltalnL > 4.0 ) { poiPlus2stdv = x ; printf(" set p2 : %d, x=%g, 2dnll=%g\n", xi, x, twoDeltalnL) ;}
if ( xi%100 == 0 ) { printf( " %4d : poi=%6.2f, 2DeltalnL = %6.2f\n", xi, x, twoDeltalnL ) ; }
}
printf("\n\n POI estimate : %g +%g -%g [%g,%g], two sigma errors: +%g -%g [%g,%g]\n\n",
poiBest,
(poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv), poiMinus1stdv, poiPlus1stdv,
(poiPlus2stdv-poiBest), (poiBest-poiMinus2stdv), poiMinus2stdv, poiPlus2stdv
) ;
printf(" %s val,pm1sig,pm2sig: %7.2f %7.2f %7.2f %7.2f %7.2f\n",
new_poi_name, poiBest, (poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv), (poiPlus2stdv-poiBest), (poiBest-poiMinus2stdv) ) ;
char htitle[1000] ;
sprintf(htitle, "%s profile likelihood scan: -2ln(L/Lm)", new_poi_name ) ;
TH1F* hscan = new TH1F("hscan", htitle, 10, poiMinVal, poiMaxVal ) ;
hscan->SetMinimum(0.) ;
hscan->SetMaximum(ymax) ;
hscan->DrawCopy() ;
graph->SetLineColor(4) ;
graph->SetLineWidth(3) ;
graph->Draw("CP") ;
gPad->SetGridx(1) ;
gPad->SetGridy(1) ;
cscan->Update() ;
TLine* line = new TLine() ;
line->SetLineColor(2) ;
line->DrawLine(poiMinVal, 1., poiPlus1stdv, 1.) ;
line->DrawLine(poiMinus1stdv,0., poiMinus1stdv, 1.) ;
line->DrawLine(poiPlus1stdv ,0., poiPlus1stdv , 1.) ;
TText* text = new TText() ;
text->SetTextSize(0.04) ;
char tstring[1000] ;
sprintf( tstring, "%s = %.1f +%.1f -%.1f", new_poi_name, poiBest, (poiPlus1stdv-poiBest), (poiBest-poiMinus1stdv) ) ;
text -> DrawTextNDC( 0.15, 0.85, tstring ) ;
sprintf( tstring, "68%% interval [%.1f, %.1f]", poiMinus1stdv, poiPlus1stdv ) ;
text -> DrawTextNDC( 0.15, 0.78, tstring ) ;
char hname[1000] ;
sprintf( hname, "hscanout_%s", new_poi_name ) ;
TH1F* hsout = new TH1F( hname,"scan results",4,0.,4.) ;
double obsVal(-1.) ;
hsout->SetBinContent(1, obsVal ) ;
hsout->SetBinContent(2, poiPlus1stdv ) ;
hsout->SetBinContent(3, poiBest ) ;
hsout->SetBinContent(4, poiMinus1stdv ) ;
TAxis* xaxis = hsout->GetXaxis() ;
xaxis->SetBinLabel(1,"Observed val.") ;
xaxis->SetBinLabel(2,"Model+1sd") ;
xaxis->SetBinLabel(3,"Model") ;
xaxis->SetBinLabel(4,"Model-1sd") ;
char outrootfile[10000] ;
sprintf( outrootfile, "%s/scan-ff-%s.root", outputdir.Data(), new_poi_name ) ;
char outpdffile[10000] ;
sprintf( outpdffile, "%s/scan-ff-%s.pdf", outputdir.Data(), new_poi_name ) ;
cscan->Update() ; cscan->Draw() ;
printf("\n Saving %s\n", outpdffile ) ;
cscan->SaveAs( outpdffile ) ;
//--- save in root file
printf("\n Saving %s\n", outrootfile ) ;
TFile fout(outrootfile,"recreate") ;
graph->Write() ;
hsout->Write() ;
fout.Close() ;
delete ws ;
wstf->Close() ;
} // constrained_scan.
// internal routine to run the inverter
HypoTestInverterResult *
RooStats::HypoTestInvTool::RunInverter(RooWorkspace * w,
const char * modelSBName, const char * modelBName,
const char * dataName, int type, int testStatType,
bool useCLs, int npoints, double poimin, double poimax,
int ntoys,
bool useNumberCounting,
const char * nuisPriorName ){
std::cout << "Running HypoTestInverter on the workspace " << w->GetName() << std::endl;
w->Print();
RooAbsData * data = w->data(dataName);
if (!data) {
Error("StandardHypoTestDemo","Not existing data %s",dataName);
return 0;
}
else
std::cout << "Using data set " << dataName << std::endl;
if (mUseVectorStore) {
RooAbsData::setDefaultStorageType(RooAbsData::Vector);
data->convertToVectorStore() ;
}
// get models from WS
// get the modelConfig out of the file
ModelConfig* bModel = (ModelConfig*) w->obj(modelBName);
ModelConfig* sbModel = (ModelConfig*) w->obj(modelSBName);
if (!sbModel) {
Error("StandardHypoTestDemo","Not existing ModelConfig %s",modelSBName);
return 0;
}
// check the model
if (!sbModel->GetPdf()) {
Error("StandardHypoTestDemo","Model %s has no pdf ",modelSBName);
return 0;
}
if (!sbModel->GetParametersOfInterest()) {
Error("StandardHypoTestDemo","Model %s has no poi ",modelSBName);
return 0;
}
if (!sbModel->GetObservables()) {
Error("StandardHypoTestInvDemo","Model %s has no observables ",modelSBName);
return 0;
}
if (!sbModel->GetSnapshot() ) {
Info("StandardHypoTestInvDemo","Model %s has no snapshot - make one using model poi",modelSBName);
sbModel->SetSnapshot( *sbModel->GetParametersOfInterest() );
}
// case of no systematics
// remove nuisance parameters from model
if (noSystematics) {
const RooArgSet * nuisPar = sbModel->GetNuisanceParameters();
if (nuisPar && nuisPar->getSize() > 0) {
std::cout << "StandardHypoTestInvDemo" << " - Switch off all systematics by setting them constant to their initial values" << std::endl;
RooStats::SetAllConstant(*nuisPar);
}
if (bModel) {
const RooArgSet * bnuisPar = bModel->GetNuisanceParameters();
if (bnuisPar)
RooStats::SetAllConstant(*bnuisPar);
}
}
if (!bModel || bModel == sbModel) {
Info("StandardHypoTestInvDemo","The background model %s does not exist",modelBName);
Info("StandardHypoTestInvDemo","Copy it from ModelConfig %s and set POI to zero",modelSBName);
bModel = (ModelConfig*) sbModel->Clone();
bModel->SetName(TString(modelSBName)+TString("_with_poi_0"));
RooRealVar * var = dynamic_cast<RooRealVar*>(bModel->GetParametersOfInterest()->first());
if (!var) return 0;
double oldval = var->getVal();
var->setVal(0);
bModel->SetSnapshot( RooArgSet(*var) );
var->setVal(oldval);
}
else {
if (!bModel->GetSnapshot() ) {
Info("StandardHypoTestInvDemo","Model %s has no snapshot - make one using model poi and 0 values ",modelBName);
RooRealVar * var = dynamic_cast<RooRealVar*>(bModel->GetParametersOfInterest()->first());
if (var) {
double oldval = var->getVal();
var->setVal(0);
bModel->SetSnapshot( RooArgSet(*var) );
var->setVal(oldval);
}
else {
Error("StandardHypoTestInvDemo","Model %s has no valid poi",modelBName);
return 0;
}
}
}
// check model has global observables when there are nuisance pdf
// for the hybrid case the globobs are not needed
if (type != 1 ) {
bool hasNuisParam = (sbModel->GetNuisanceParameters() && sbModel->GetNuisanceParameters()->getSize() > 0);
bool hasGlobalObs = (sbModel->GetGlobalObservables() && sbModel->GetGlobalObservables()->getSize() > 0);
if (hasNuisParam && !hasGlobalObs ) {
// try to see if model has nuisance parameters first
RooAbsPdf * constrPdf = RooStats::MakeNuisancePdf(*sbModel,"nuisanceConstraintPdf_sbmodel");
if (constrPdf) {
Warning("StandardHypoTestInvDemo","Model %s has nuisance parameters but no global observables associated",sbModel->GetName());
Warning("StandardHypoTestInvDemo","\tThe effect of the nuisance parameters will not be treated correctly ");
}
}
}
// run first a data fit
const RooArgSet * poiSet = sbModel->GetParametersOfInterest();
RooRealVar *poi = (RooRealVar*)poiSet->first();
std::cout << "StandardHypoTestInvDemo : POI initial value: " << poi->GetName() << " = " << poi->getVal() << std::endl;
// fit the data first (need to use constraint )
TStopwatch tw;
bool doFit = initialFit;
if (testStatType == 0 && initialFit == -1) doFit = false; // case of LEP test statistic
if (type == 3 && initialFit == -1) doFit = false; // case of Asymptoticcalculator with nominal Asimov
double poihat = 0;
if (minimizerType.size()==0) minimizerType = ROOT::Math::MinimizerOptions::DefaultMinimizerType();
else
ROOT::Math::MinimizerOptions::SetDefaultMinimizer(minimizerType.c_str());
Info("StandardHypoTestInvDemo","Using %s as minimizer for computing the test statistic",
ROOT::Math::MinimizerOptions::DefaultMinimizerType().c_str() );
if (doFit) {
// do the fit : By doing a fit the POI snapshot (for S+B) is set to the fit value
// and the nuisance parameters nominal values will be set to the fit value.
// This is relevant when using LEP test statistics
Info( "StandardHypoTestInvDemo"," Doing a first fit to the observed data ");
RooArgSet constrainParams;
if (sbModel->GetNuisanceParameters() ) constrainParams.add(*sbModel->GetNuisanceParameters());
RooStats::RemoveConstantParameters(&constrainParams);
tw.Start();
RooFitResult * fitres = sbModel->GetPdf()->fitTo(*data,InitialHesse(false), Hesse(false),
Minimizer(minimizerType.c_str(),"Migrad"), Strategy(0), PrintLevel(mPrintLevel), Constrain(constrainParams), Save(true) );
if (fitres->status() != 0) {
Warning("StandardHypoTestInvDemo","Fit to the model failed - try with strategy 1 and perform first an Hesse computation");
fitres = sbModel->GetPdf()->fitTo(*data,InitialHesse(true), Hesse(false),Minimizer(minimizerType.c_str(),"Migrad"), Strategy(1), PrintLevel(mPrintLevel+1), Constrain(constrainParams), Save(true) );
}
if (fitres->status() != 0)
Warning("StandardHypoTestInvDemo"," Fit still failed - continue anyway.....");
poihat = poi->getVal();
std::cout << "StandardHypoTestInvDemo - Best Fit value : " << poi->GetName() << " = "
<< poihat << " +/- " << poi->getError() << std::endl;
std::cout << "Time for fitting : "; tw.Print();
//save best fit value in the poi snapshot
sbModel->SetSnapshot(*sbModel->GetParametersOfInterest());
std::cout << "StandardHypoTestInvo: snapshot of S+B Model " << sbModel->GetName()
<< " is set to the best fit value" << std::endl;
}
// print a message in case of LEP test statistics because it affects result by doing or not doing a fit
if (testStatType == 0) {
if (!doFit)
Info("StandardHypoTestInvDemo","Using LEP test statistic - an initial fit is not done and the TS will use the nuisances at the model value");
else
Info("StandardHypoTestInvDemo","Using LEP test statistic - an initial fit has been done and the TS will use the nuisances at the best fit value");
}
// build test statistics and hypotest calculators for running the inverter
SimpleLikelihoodRatioTestStat slrts(*sbModel->GetPdf(),*bModel->GetPdf());
// null parameters must includes snapshot of poi plus the nuisance values
RooArgSet nullParams(*sbModel->GetSnapshot());
if (sbModel->GetNuisanceParameters()) nullParams.add(*sbModel->GetNuisanceParameters());
if (sbModel->GetSnapshot()) slrts.SetNullParameters(nullParams);
RooArgSet altParams(*bModel->GetSnapshot());
if (bModel->GetNuisanceParameters()) altParams.add(*bModel->GetNuisanceParameters());
if (bModel->GetSnapshot()) slrts.SetAltParameters(altParams);
// ratio of profile likelihood - need to pass snapshot for the alt
RatioOfProfiledLikelihoodsTestStat
ropl(*sbModel->GetPdf(), *bModel->GetPdf(), bModel->GetSnapshot());
ropl.SetSubtractMLE(false);
if (testStatType == 11) ropl.SetSubtractMLE(true);
ropl.SetPrintLevel(mPrintLevel);
ropl.SetMinimizer(minimizerType.c_str());
ProfileLikelihoodTestStat profll(*sbModel->GetPdf());
if (testStatType == 3) profll.SetOneSided(true);
if (testStatType == 4) profll.SetSigned(true);
profll.SetMinimizer(minimizerType.c_str());
profll.SetPrintLevel(mPrintLevel);
profll.SetReuseNLL(mOptimize);
slrts.SetReuseNLL(mOptimize);
ropl.SetReuseNLL(mOptimize);
if (mOptimize) {
profll.SetStrategy(0);
ropl.SetStrategy(0);
ROOT::Math::MinimizerOptions::SetDefaultStrategy(0);
}
if (mMaxPoi > 0) poi->setMax(mMaxPoi); // increase limit
MaxLikelihoodEstimateTestStat maxll(*sbModel->GetPdf(),*poi);
NumEventsTestStat nevtts;
AsymptoticCalculator::SetPrintLevel(mPrintLevel);
// create the HypoTest calculator class
HypoTestCalculatorGeneric * hc = 0;
if (type == 0) hc = new FrequentistCalculator(*data, *bModel, *sbModel);
else if (type == 1) hc = new HybridCalculator(*data, *bModel, *sbModel);
// else if (type == 2 ) hc = new AsymptoticCalculator(*data, *bModel, *sbModel, false, mAsimovBins);
// else if (type == 3 ) hc = new AsymptoticCalculator(*data, *bModel, *sbModel, true, mAsimovBins); // for using Asimov data generated with nominal values
else if (type == 2 ) hc = new AsymptoticCalculator(*data, *bModel, *sbModel, false );
else if (type == 3 ) hc = new AsymptoticCalculator(*data, *bModel, *sbModel, true ); // for using Asimov data generated with nominal values
else {
Error("StandardHypoTestInvDemo","Invalid - calculator type = %d supported values are only :\n\t\t\t 0 (Frequentist) , 1 (Hybrid) , 2 (Asymptotic) ",type);
return 0;
}
// set the test statistic
TestStatistic * testStat = 0;
if (testStatType == 0) testStat = &slrts;
if (testStatType == 1 || testStatType == 11) testStat = &ropl;
if (testStatType == 2 || testStatType == 3 || testStatType == 4) testStat = &profll;
if (testStatType == 5) testStat = &maxll;
if (testStatType == 6) testStat = &nevtts;
if (testStat == 0) {
Error("StandardHypoTestInvDemo","Invalid - test statistic type = %d supported values are only :\n\t\t\t 0 (SLR) , 1 (Tevatron) , 2 (PLR), 3 (PLR1), 4(MLE)",testStatType);
return 0;
}
ToyMCSampler *toymcs = (ToyMCSampler*)hc->GetTestStatSampler();
if (toymcs && (type == 0 || type == 1) ) {
// look if pdf is number counting or extended
if (sbModel->GetPdf()->canBeExtended() ) {
if (useNumberCounting) Warning("StandardHypoTestInvDemo","Pdf is extended: but number counting flag is set: ignore it ");
}
else {
// for not extended pdf
if (!useNumberCounting ) {
int nEvents = data->numEntries();
Info("StandardHypoTestInvDemo","Pdf is not extended: number of events to generate taken from observed data set is %d",nEvents);
toymcs->SetNEventsPerToy(nEvents);
}
else {
Info("StandardHypoTestInvDemo","using a number counting pdf");
toymcs->SetNEventsPerToy(1);
}
}
toymcs->SetTestStatistic(testStat);
if (data->isWeighted() && !mGenerateBinned) {
Info("StandardHypoTestInvDemo","Data set is weighted, nentries = %d and sum of weights = %8.1f but toy generation is unbinned - it would be faster to set mGenerateBinned to true\n",data->numEntries(), data->sumEntries());
}
toymcs->SetGenerateBinned(mGenerateBinned);
toymcs->SetUseMultiGen(mOptimize);
if (mGenerateBinned && sbModel->GetObservables()->getSize() > 2) {
Warning("StandardHypoTestInvDemo","generate binned is activated but the number of ovservable is %d. Too much memory could be needed for allocating all the bins",sbModel->GetObservables()->getSize() );
}
// set the random seed if needed
if (mRandomSeed >= 0) RooRandom::randomGenerator()->SetSeed(mRandomSeed);
}
// specify if need to re-use same toys
if (reuseAltToys) {
hc->UseSameAltToys();
}
if (type == 1) {
HybridCalculator *hhc = dynamic_cast<HybridCalculator*> (hc);
assert(hhc);
hhc->SetToys(ntoys,ntoys/mNToysRatio); // can use less ntoys for b hypothesis
// remove global observables from ModelConfig (this is probably not needed anymore in 5.32)
bModel->SetGlobalObservables(RooArgSet() );
sbModel->SetGlobalObservables(RooArgSet() );
// check for nuisance prior pdf in case of nuisance parameters
if (bModel->GetNuisanceParameters() || sbModel->GetNuisanceParameters() ) {
// fix for using multigen (does not work in this case)
toymcs->SetUseMultiGen(false);
ToyMCSampler::SetAlwaysUseMultiGen(false);
RooAbsPdf * nuisPdf = 0;
if (nuisPriorName) nuisPdf = w->pdf(nuisPriorName);
// use prior defined first in bModel (then in SbModel)
if (!nuisPdf) {
Info("StandardHypoTestInvDemo","No nuisance pdf given for the HybridCalculator - try to deduce pdf from the model");
if (bModel->GetPdf() && bModel->GetObservables() )
nuisPdf = RooStats::MakeNuisancePdf(*bModel,"nuisancePdf_bmodel");
else
nuisPdf = RooStats::MakeNuisancePdf(*sbModel,"nuisancePdf_sbmodel");
}
if (!nuisPdf ) {
if (bModel->GetPriorPdf()) {
nuisPdf = bModel->GetPriorPdf();
Info("StandardHypoTestInvDemo","No nuisance pdf given - try to use %s that is defined as a prior pdf in the B model",nuisPdf->GetName());
}
else {
Error("StandardHypoTestInvDemo","Cannnot run Hybrid calculator because no prior on the nuisance parameter is specified or can be derived");
return 0;
}
}
assert(nuisPdf);
Info("StandardHypoTestInvDemo","Using as nuisance Pdf ... " );
nuisPdf->Print();
const RooArgSet * nuisParams = (bModel->GetNuisanceParameters() ) ? bModel->GetNuisanceParameters() : sbModel->GetNuisanceParameters();
RooArgSet * np = nuisPdf->getObservables(*nuisParams);
if (np->getSize() == 0) {
Warning("StandardHypoTestInvDemo","Prior nuisance does not depend on nuisance parameters. They will be smeared in their full range");
}
delete np;
hhc->ForcePriorNuisanceAlt(*nuisPdf);
hhc->ForcePriorNuisanceNull(*nuisPdf);
}
}
else if (type == 2 || type == 3) {
if (testStatType == 3) ((AsymptoticCalculator*) hc)->SetOneSided(true);
if (testStatType != 2 && testStatType != 3)
Warning("StandardHypoTestInvDemo","Only the PL test statistic can be used with AsymptoticCalculator - use by default a two-sided PL");
}
else if (type == 0 || type == 1)
((FrequentistCalculator*) hc)->SetToys(ntoys,ntoys/mNToysRatio);
// Get the result
RooMsgService::instance().getStream(1).removeTopic(RooFit::NumIntegration);
HypoTestInverter calc(*hc);
calc.SetConfidenceLevel(0.95);
calc.UseCLs(useCLs);
calc.SetVerbose(true);
// can speed up using proof-lite
if (mUseProof && mNWorkers > 1) {
ProofConfig pc(*w, mNWorkers, "", kFALSE);
toymcs->SetProofConfig(&pc); // enable proof
}
if (npoints > 0) {
if (poimin > poimax) {
// if no min/max given scan between MLE and +4 sigma
poimin = int(poihat);
poimax = int(poihat + 4 * poi->getError());
}
std::cout << "Doing a fixed scan in interval : " << poimin << " , " << poimax << std::endl;
calc.SetFixedScan(npoints,poimin,poimax);
}
else {
//poi->setMax(10*int( (poihat+ 10 *poi->getError() )/10 ) );
std::cout << "Doing an automatic scan in interval : " << poi->getMin() << " , " << poi->getMax() << std::endl;
}
tw.Start();
HypoTestInverterResult * r = calc.GetInterval();
std::cout << "Time to perform limit scan \n";
tw.Print();
if (mRebuild) {
calc.SetCloseProof(1);
tw.Start();
SamplingDistribution * limDist = calc.GetUpperLimitDistribution(true,mNToyToRebuild);
std::cout << "Time to rebuild distributions " << std::endl;
tw.Print();
if (limDist) {
std::cout << "expected up limit " << limDist->InverseCDF(0.5) << " +/- "
<< limDist->InverseCDF(0.16) << " "
<< limDist->InverseCDF(0.84) << "\n";
//update r to a new updated result object containing the rebuilt expected p-values distributions
// (it will not recompute the expected limit)
if (r) delete r; // need to delete previous object since GetInterval will return a cloned copy
r = calc.GetInterval();
}
else
std::cout << "ERROR : failed to re-build distributions " << std::endl;
}
return r;
}
void printCtauFrom2DParameters(RooWorkspace myws, TPad* Pad, bool isPbPb, string pdfName, bool isWeighted)
{
Pad->cd();
TLatex *t = new TLatex(); t->SetNDC(); t->SetTextSize(0.026); float dy = 0.045;
RooArgSet* Parameters = (RooArgSet*)myws.pdf(pdfName.c_str())->getParameters(RooArgSet(*myws.var("invMass"), *myws.var("ctau"), *myws.var("ctauErr")))->selectByAttrib("Constant",kFALSE);
TIterator* parIt = Parameters->createIterator();
for (RooRealVar* it = (RooRealVar*)parIt->Next(); it!=NULL; it = (RooRealVar*)parIt->Next() ) {
stringstream ss(it->GetName()); string s1, s2, s3, label;
getline(ss, s1, '_'); getline(ss, s2, '_'); getline(ss, s3, '_');
// Parse the parameter's labels
if(s1=="invMass"){continue;} else if(s1=="ctau"){continue;} else if(s1=="MassRatio"){continue;}
else if(s1=="One"){continue;} else if(s1=="mMin"){continue;} else if(s1=="mMax"){continue;}
else if(s1=="rSigma21"){ s1="(#sigma_{2}/#sigma_{1})"; }
else if(s1=="rSigma32"){ s1="(#sigma_{3}/#sigma_{2})"; }
else if(s1=="rSigma43"){ s1="(#sigma_{4}/#sigma_{3})"; }
else if(s1.find("sigma")!=std::string::npos || s1.find("lambda")!=std::string::npos || s1.find("alpha")!=std::string::npos){
s1=Form("#%s",s1.c_str());
}
if(s2=="CtauRes") { s2="Res"; }
else if(s2=="JpsiNoPR") { s2="J/#psi[NoPR]"; }
else if(s2=="JpsiPR") { s2="J/#psi[PR]"; }
else if(s2=="Jpsi" && (s1=="N" || s1=="b")) { s2="J/#psi"; }
else if(s2=="Psi2SNoPR") { s2="#psi(2S)[NoPR]"; }
else if(s2=="Psi2SPR") { s2="#psi(2S)[PR]"; }
else if(s2=="Psi2S" && (s1=="N" || s1=="b")) { s2="#psi(2S)"; }
else if(s2=="BkgNoPR") { s2="bkg[NoPR]"; }
else if(s2=="BkgPR") { s2="bkg[PR]"; }
else if(s2=="Bkg" && (s1=="N" || s1=="b")) { s2="bkg"; }
else {continue;}
if(s3!=""){
label=Form("%s_{%s}^{%s}", s1.c_str(), s2.c_str(), s3.c_str());
}
else {
label=Form("%s^{%s}", s1.c_str(), s2.c_str());
}
// Print the parameter's results
if(s1=="N"){
t->DrawLatex(0.69, 0.75-dy, Form((isWeighted?"%s = %.6f#pm%.6f ":"%s = %.0f#pm%.0f "), label.c_str(), it->getValV(), it->getError())); dy+=0.045;
}
else if(s1=="b"){
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.4f#pm%.4f ", label.c_str(), it->getValV(), it->getError())); dy+=0.045;
}
else if(s1.find("sigma")!=std::string::npos) {
if (s2.find("Res")!=std::string::npos) {
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.2f#pm%.2f", label.c_str(), it->getValV(), it->getError())); dy+=0.045;
} else {
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.2f#pm%.2f mm", label.c_str(), it->getValV(), it->getError())); dy+=0.045;
}
}
else if(s1.find("lambda")!=std::string::npos){
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.4f#pm%.4f", label.c_str(), it->getValV(), it->getError())); dy+=0.045;
}
else if(s1.find("ctau")!=std::string::npos){
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.4f#pm%.4f mm", (label.insert(1, string("#"))).c_str(), it->getValV(), it->getError())); dy+=0.045;
}
else {
t->DrawLatex(0.69, 0.75-dy, Form("%s = %.4f#pm%.4f", label.c_str(), it->getValV(), it->getError())); dy+=0.045;
}
}
};
void StandardBayesianNumericalDemo(const char* infile = "",
const char* workspaceName = "combined",
const char* modelConfigName = "ModelConfig",
const char* dataName = "obsData") {
// option definitions
double confLevel = optBayes.confLevel;
TString integrationType = optBayes.integrationType;
int nToys = optBayes.nToys;
bool scanPosterior = optBayes.scanPosterior;
int nScanPoints = optBayes.nScanPoints;
int intervalType = optBayes.intervalType;
int maxPOI = optBayes.maxPOI;
double nSigmaNuisance = optBayes.nSigmaNuisance;
/////////////////////////////////////////////////////////////
// First part is just to access a user-defined file
// or create the standard example file if it doesn't exist
////////////////////////////////////////////////////////////
const char* filename = "";
if (!strcmp(infile,"")) {
filename = "results/example_combined_GaussExample_model.root";
bool fileExist = !gSystem->AccessPathName(filename); // note opposite return code
// if file does not exists generate with histfactory
if (!fileExist) {
#ifdef _WIN32
cout << "HistFactory file cannot be generated on Windows - exit" << endl;
return;
#endif
// Normally this would be run on the command line
cout <<"will run standard hist2workspace example"<<endl;
gROOT->ProcessLine(".! prepareHistFactory .");
gROOT->ProcessLine(".! hist2workspace config/example.xml");
cout <<"\n\n---------------------"<<endl;
cout <<"Done creating example input"<<endl;
cout <<"---------------------\n\n"<<endl;
}
}
else
filename = infile;
// Try to open the file
TFile *file = TFile::Open(filename);
// if input file was specified byt not found, quit
if(!file ){
cout <<"StandardRooStatsDemoMacro: Input file " << filename << " is not found" << endl;
return;
}
/////////////////////////////////////////////////////////////
// Tutorial starts here
////////////////////////////////////////////////////////////
// get the workspace out of the file
RooWorkspace* w = (RooWorkspace*) file->Get(workspaceName);
if(!w){
cout <<"workspace not found" << endl;
return;
}
// get the modelConfig out of the file
ModelConfig* mc = (ModelConfig*) w->obj(modelConfigName);
// get the modelConfig out of the file
RooAbsData* data = w->data(dataName);
// make sure ingredients are found
if(!data || !mc){
w->Print();
cout << "data or ModelConfig was not found" <<endl;
return;
}
/////////////////////////////////////////////
// create and use the BayesianCalculator
// to find and plot the 95% credible interval
// on the parameter of interest as specified
// in the model config
// before we do that, we must specify our prior
// it belongs in the model config, but it may not have
// been specified
RooUniform prior("prior","",*mc->GetParametersOfInterest());
w->import(prior);
mc->SetPriorPdf(*w->pdf("prior"));
// do without systematics
//mc->SetNuisanceParameters(RooArgSet() );
if (nSigmaNuisance > 0) {
RooAbsPdf * pdf = mc->GetPdf();
assert(pdf);
RooFitResult * res = pdf->fitTo(*data, Save(true), Minimizer(ROOT::Math::MinimizerOptions::DefaultMinimizerType().c_str()), Hesse(true),
PrintLevel(ROOT::Math::MinimizerOptions::DefaultPrintLevel()-1) );
res->Print();
RooArgList nuisPar(*mc->GetNuisanceParameters());
for (int i = 0; i < nuisPar.getSize(); ++i) {
RooRealVar * v = dynamic_cast<RooRealVar*> (&nuisPar[i] );
assert( v);
v->setMin( TMath::Max( v->getMin(), v->getVal() - nSigmaNuisance * v->getError() ) );
v->setMax( TMath::Min( v->getMax(), v->getVal() + nSigmaNuisance * v->getError() ) );
std::cout << "setting interval for nuisance " << v->GetName() << " : [ " << v->getMin() << " , " << v->getMax() << " ]" << std::endl;
}
}
BayesianCalculator bayesianCalc(*data,*mc);
bayesianCalc.SetConfidenceLevel(confLevel); // 95% interval
// default of the calculator is central interval. here use shortest , central or upper limit depending on input
// doing a shortest interval might require a longer time since it requires a scan of the posterior function
if (intervalType == 0) bayesianCalc.SetShortestInterval(); // for shortest interval
if (intervalType == 1) bayesianCalc.SetLeftSideTailFraction(0.5); // for central interval
if (intervalType == 2) bayesianCalc.SetLeftSideTailFraction(0.); // for upper limit
if (!integrationType.IsNull() ) {
bayesianCalc.SetIntegrationType(integrationType); // set integrationType
bayesianCalc.SetNumIters(nToys); // set number of ietrations (i.e. number of toys for MC integrations)
}
// in case of toyMC make a nnuisance pdf
if (integrationType.Contains("TOYMC") ) {
RooAbsPdf * nuisPdf = RooStats::MakeNuisancePdf(*mc, "nuisance_pdf");
cout << "using TOYMC integration: make nuisance pdf from the model " << std::endl;
nuisPdf->Print();
bayesianCalc.ForceNuisancePdf(*nuisPdf);
scanPosterior = true; // for ToyMC the posterior is scanned anyway so used given points
}
// compute interval by scanning the posterior function
if (scanPosterior)
bayesianCalc.SetScanOfPosterior(nScanPoints);
RooRealVar* poi = (RooRealVar*) mc->GetParametersOfInterest()->first();
if (maxPOI != -999 && maxPOI > poi->getMin())
poi->setMax(maxPOI);
SimpleInterval* interval = bayesianCalc.GetInterval();
// print out the iterval on the first Parameter of Interest
cout << "\n>>>> RESULT : " << confLevel*100 << "% interval on " << poi->GetName()<<" is : ["<<
interval->LowerLimit() << ", "<<
interval->UpperLimit() <<"] "<<endl;
// make a plot
// since plotting may take a long time (it requires evaluating
// the posterior in many points) this command will speed up
// by reducing the number of points to plot - do 50
// ignore errors of PDF if is zero
RooAbsReal::setEvalErrorLoggingMode(RooAbsReal::Ignore) ;
cout << "\nDrawing plot of posterior function....." << endl;
// always plot using numer of scan points
bayesianCalc.SetScanOfPosterior(nScanPoints);
RooPlot * plot = bayesianCalc.GetPosteriorPlot();
plot->Draw();
}
void ws_cls_hybrid1_ag( const char* wsfile = "output-files/expected-ws-lm9-2BL.root", bool isBgonlyStudy=false, double poiVal = 150.0, int nToys=100, bool makeTtree=true, int verbLevel=0 ) {
TTree* toytt(0x0) ;
TFile* ttfile(0x0) ;
int tt_gen_Nsig ;
int tt_gen_Nsb ;
int tt_gen_Nsig_sl ;
int tt_gen_Nsb_sl ;
int tt_gen_Nsig_ldp ;
int tt_gen_Nsb_ldp ;
int tt_gen_Nsig_ee ;
int tt_gen_Nsb_ee ;
int tt_gen_Nsig_mm ;
int tt_gen_Nsb_mm ;
double tt_testStat ;
double tt_dataTestStat ;
double tt_hypo_mu_susy_sig ;
char ttname[1000] ;
char tttitle[1000] ;
if ( makeTtree ) {
ttfile = gDirectory->GetFile() ;
if ( ttfile == 0x0 ) { printf("\n\n\n *** asked for a ttree but no open file???\n\n") ; return ; }
if ( isBgonlyStudy ) {
sprintf( ttname, "toytt_%.0f_bgo", poiVal ) ;
sprintf( tttitle, "Toy study for background only, mu_susy_sig = %.0f", poiVal ) ;
} else {
sprintf( ttname, "toytt_%.0f_spb", poiVal ) ;
sprintf( tttitle, "Toy study for signal+background, mu_susy_sig = %.0f", poiVal ) ;
}
printf("\n\n Creating TTree : %s : %s\n\n", ttname, tttitle ) ;
gDirectory->pwd() ;
gDirectory->ls() ;
toytt = new TTree( ttname, tttitle ) ;
gDirectory->ls() ;
toytt -> Branch( "gen_Nsig" , &tt_gen_Nsig , "gen_Nsig/I" ) ;
toytt -> Branch( "gen_Nsb" , &tt_gen_Nsb , "gen_Nsb/I" ) ;
toytt -> Branch( "gen_Nsig_sl" , &tt_gen_Nsig_sl , "gen_Nsig_sl/I" ) ;
toytt -> Branch( "gen_Nsb_sl" , &tt_gen_Nsb_sl , "gen_Nsb_sl/I" ) ;
toytt -> Branch( "gen_Nsig_ldp" , &tt_gen_Nsig_ldp , "gen_Nsig_ldp/I" ) ;
toytt -> Branch( "gen_Nsb_ldp" , &tt_gen_Nsb_ldp , "gen_Nsb_ldp/I" ) ;
toytt -> Branch( "gen_Nsig_ee" , &tt_gen_Nsig_ee , "gen_Nsig_ee/I" ) ;
toytt -> Branch( "gen_Nsb_ee" , &tt_gen_Nsb_ee , "gen_Nsb_ee/I" ) ;
toytt -> Branch( "gen_Nsig_mm" , &tt_gen_Nsig_mm , "gen_Nsig_mm/I" ) ;
toytt -> Branch( "gen_Nsb_mm" , &tt_gen_Nsb_mm , "gen_Nsb_mm/I" ) ;
toytt -> Branch( "testStat" , &tt_testStat , "testStat/D" ) ;
toytt -> Branch( "dataTestStat" , &tt_dataTestStat , "dataTestStat/D" ) ;
toytt -> Branch( "hypo_mu_susy_sig" , &tt_hypo_mu_susy_sig , "hypo_mu_susy_sig/D" ) ;
}
//--- Tell RooFit to shut up about anything less important than an ERROR.
RooMsgService::instance().setGlobalKillBelow(RooFit::ERROR) ;
random_ng = new TRandom2(12345) ;
/// char sel[100] ;
/// if ( strstr( wsfile, "1BL" ) != 0 ) {
/// sprintf( sel, "1BL" ) ;
/// } else if ( strstr( wsfile, "2BL" ) != 0 ) {
/// sprintf( sel, "2BL" ) ;
/// } else if ( strstr( wsfile, "3B" ) != 0 ) {
/// sprintf( sel, "3B" ) ;
/// } else if ( strstr( wsfile, "1BT" ) != 0 ) {
/// sprintf( sel, "1BT" ) ;
/// } else if ( strstr( wsfile, "2BT" ) != 0 ) {
/// sprintf( sel, "2BT" ) ;
/// } else {
/// printf("\n\n\n *** can't figure out which selection this is. I quit.\n\n" ) ;
/// return ;
/// }
/// printf("\n\n selection is %s\n\n", sel ) ;
TFile* wstf = new TFile( wsfile ) ;
RooWorkspace* ws = dynamic_cast<RooWorkspace*>( wstf->Get("ws") );
ws->Print() ;
RooDataSet* rds = (RooDataSet*) ws->obj( "ra2b_observed_rds" ) ;
printf("\n\n\n ===== RooDataSet ====================\n\n") ;
rds->Print() ;
rds->printMultiline(cout, 1, kTRUE, "") ;
ModelConfig* modelConfig = (ModelConfig*) ws->obj( "SbModel" ) ;
RooAbsPdf* likelihood = modelConfig->GetPdf() ;
const RooArgSet* nuisanceParameters = modelConfig->GetNuisanceParameters() ;
RooRealVar* rrv_mu_susy_sig = ws->var("mu_susy_sig") ;
if ( rrv_mu_susy_sig == 0x0 ) {
printf("\n\n\n *** can't find mu_susy_sig in workspace. Quitting.\n\n\n") ;
return ;
}
//// printf("\n\n\n ===== Doing a fit ====================\n\n") ;
//// RooFitResult* preFitResult = likelihood->fitTo( *rds, Save(true) ) ;
//// const RooArgList preFitFloatVals = preFitResult->floatParsFinal() ;
//// {
//// TIterator* parIter = preFitFloatVals.createIterator() ;
//// while ( RooRealVar* par = (RooRealVar*) parIter->Next() ) {
//// printf(" %20s : %8.2f\n", par->GetName(), par->getVal() ) ;
//// }
//// }
//--- Get pointers to the model predictions of the observables.
rfv_n_sig = ws->function("n_sig") ;
rfv_n_sb = ws->function("n_sb") ;
rfv_n_sig_sl = ws->function("n_sig_sl") ;
rfv_n_sb_sl = ws->function("n_sb_sl") ;
rfv_n_sig_ldp = ws->function("n_sig_ldp") ;
rfv_n_sb_ldp = ws->function("n_sb_ldp") ;
rfv_n_sig_ee = ws->function("n_sig_ee") ;
rfv_n_sb_ee = ws->function("n_sb_ee") ;
rfv_n_sig_mm = ws->function("n_sig_mm") ;
rfv_n_sb_mm = ws->function("n_sb_mm") ;
if ( rfv_n_sig == 0x0 ) { printf("\n\n\n *** can't find n_sig in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sb == 0x0 ) { printf("\n\n\n *** can't find n_sb in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sig_sl == 0x0 ) { printf("\n\n\n *** can't find n_sig_sl in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sb_sl == 0x0 ) { printf("\n\n\n *** can't find n_sb_sl in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sig_ldp == 0x0 ) { printf("\n\n\n *** can't find n_sig_ldp in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sb_ldp == 0x0 ) { printf("\n\n\n *** can't find n_sb_ldp in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sig_ee == 0x0 ) { printf("\n\n\n *** can't find n_sig_ee in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sb_ee == 0x0 ) { printf("\n\n\n *** can't find n_sb_ee in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sig_mm == 0x0 ) { printf("\n\n\n *** can't find n_sig_mm in workspace. Quitting.\n\n\n") ; return ; }
if ( rfv_n_sb_mm == 0x0 ) { printf("\n\n\n *** can't find n_sb_mm in workspace. Quitting.\n\n\n") ; return ; }
//--- Get pointers to the observables.
const RooArgSet* dsras = rds->get() ;
TIterator* obsIter = dsras->createIterator() ;
while ( RooRealVar* obs = (RooRealVar*) obsIter->Next() ) {
if ( strcmp( obs->GetName(), "Nsig" ) == 0 ) { rrv_Nsig = obs ; }
if ( strcmp( obs->GetName(), "Nsb" ) == 0 ) { rrv_Nsb = obs ; }
if ( strcmp( obs->GetName(), "Nsig_sl" ) == 0 ) { rrv_Nsig_sl = obs ; }
if ( strcmp( obs->GetName(), "Nsb_sl" ) == 0 ) { rrv_Nsb_sl = obs ; }
if ( strcmp( obs->GetName(), "Nsig_ldp" ) == 0 ) { rrv_Nsig_ldp = obs ; }
if ( strcmp( obs->GetName(), "Nsb_ldp" ) == 0 ) { rrv_Nsb_ldp = obs ; }
if ( strcmp( obs->GetName(), "Nsig_ee" ) == 0 ) { rrv_Nsig_ee = obs ; }
if ( strcmp( obs->GetName(), "Nsb_ee" ) == 0 ) { rrv_Nsb_ee = obs ; }
if ( strcmp( obs->GetName(), "Nsig_mm" ) == 0 ) { rrv_Nsig_mm = obs ; }
if ( strcmp( obs->GetName(), "Nsb_mm" ) == 0 ) { rrv_Nsb_mm = obs ; }
}
if ( rrv_Nsig == 0x0 ) { printf("\n\n\n *** can't find Nsig in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsb == 0x0 ) { printf("\n\n\n *** can't find Nsb in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsig_sl == 0x0 ) { printf("\n\n\n *** can't find Nsig_sl in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsb_sl == 0x0 ) { printf("\n\n\n *** can't find Nsb_sl in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsig_ldp == 0x0 ) { printf("\n\n\n *** can't find Nsig_ldp in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsb_ldp == 0x0 ) { printf("\n\n\n *** can't find Nsb_ldp in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsig_ee == 0x0 ) { printf("\n\n\n *** can't find Nsig_ee in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsb_ee == 0x0 ) { printf("\n\n\n *** can't find Nsb_ee in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsig_mm == 0x0 ) { printf("\n\n\n *** can't find Nsig_mm in dataset. Quitting.\n\n\n") ; return ; }
if ( rrv_Nsb_mm == 0x0 ) { printf("\n\n\n *** can't find Nsb_mm in dataset. Quitting.\n\n\n") ; return ; }
printf("\n\n\n === Model values for observables\n\n") ;
printObservables() ;
//--- save the actual values of the observables.
saveObservables() ;
//--- evaluate the test stat on the data: fit with susy floating.
rrv_mu_susy_sig->setVal( poiVal ) ;
rrv_mu_susy_sig->setConstant( kTRUE ) ;
printf("\n\n\n ====== Fitting the data with susy fixed.\n\n") ;
RooFitResult* dataFitResultSusyFixed = likelihood->fitTo(*rds, Save(true));
int dataSusyFixedFitCovQual = dataFitResultSusyFixed->covQual() ;
if ( dataSusyFixedFitCovQual != 3 ) { printf("\n\n\n *** Failed fit! Cov qual %d. Quitting.\n\n", dataSusyFixedFitCovQual ) ; return ; }
double dataFitSusyFixedNll = dataFitResultSusyFixed->minNll() ;
rrv_mu_susy_sig->setVal( 0.0 ) ;
rrv_mu_susy_sig->setConstant( kFALSE ) ;
printf("\n\n\n ====== Fitting the data with susy floating.\n\n") ;
RooFitResult* dataFitResultSusyFloat = likelihood->fitTo(*rds, Save(true));
int dataSusyFloatFitCovQual = dataFitResultSusyFloat->covQual() ;
if ( dataSusyFloatFitCovQual != 3 ) { printf("\n\n\n *** Failed fit! Cov qual %d. Quitting.\n\n", dataSusyFloatFitCovQual ) ; return ; }
double dataFitSusyFloatNll = dataFitResultSusyFloat->minNll() ;
double dataTestStat = 2.*( dataFitSusyFixedNll - dataFitSusyFloatNll) ;
printf("\n\n\n Data value of test stat : %8.2f\n", dataTestStat ) ;
printf("\n\n\n === Nuisance parameters\n\n") ;
{
int npi(0) ;
TIterator* npIter = nuisanceParameters->createIterator() ;
while ( RooRealVar* np_rrv = (RooRealVar*) npIter->Next() ) {
np_initial_val[npi] = np_rrv->getVal() ; //--- I am assuming that the order of the NPs in the iterator does not change.
TString npname( np_rrv->GetName() ) ;
npname.ReplaceAll("_prim","") ;
RooAbsReal* np_rfv = ws->function( npname ) ;
TString pdfname( np_rrv->GetName() ) ;
pdfname.ReplaceAll("_prim","") ;
pdfname.Prepend("pdf_") ;
RooAbsPdf* np_pdf = ws->pdf( pdfname ) ;
if ( np_pdf == 0x0 ) { printf("\n\n *** Can't find nuisance parameter pdf with name %s.\n\n", pdfname.Data() ) ; }
if ( np_rfv != 0x0 ) {
printf(" %20s : %8.2f , %20s, %8.2f\n", np_rrv->GetName(), np_rrv->getVal(), np_rfv->GetName(), np_rfv->getVal() ) ;
} else {
printf(" %20s : %8.2f\n", np_rrv->GetName(), np_rrv->getVal() ) ;
}
npi++ ;
} // np_rrv iterator.
np_count = npi ;
}
tt_dataTestStat = dataTestStat ;
tt_hypo_mu_susy_sig = poiVal ;
printf("\n\n\n === Doing the toys\n\n") ;
int nToyOK(0) ;
int nToyWorseThanData(0) ;
for ( int ti=0; ti<nToys; ti++ ) {
printf("\n\n\n ======= Toy %4d\n\n\n", ti ) ;
//--- 1) pick values for the nuisance parameters from the PDFs and fix them.
{
TIterator* npIter = nuisanceParameters->createIterator() ;
while ( RooRealVar* np_rrv = (RooRealVar*) npIter->Next() ) {
TString pdfname( np_rrv->GetName() ) ;
pdfname.ReplaceAll("_prim","") ;
pdfname.Prepend("pdf_") ;
RooAbsPdf* np_pdf = ws->pdf( pdfname ) ;
if ( np_pdf == 0x0 ) { printf("\n\n *** Can't find nuisance parameter pdf with name %s.\n\n", pdfname.Data() ) ; return ; }
RooDataSet* nprds = np_pdf->generate( RooArgSet(*np_rrv) ,1) ;
const RooArgSet* npdsras = nprds->get() ;
TIterator* valIter = npdsras->createIterator() ;
RooRealVar* val = (RooRealVar*) valIter->Next() ;
//--- reset the value of the nuisance parameter and fix it for the toy model definition fit.
np_rrv->setVal( val->getVal() ) ;
np_rrv->setConstant( kTRUE ) ;
TString npname( np_rrv->GetName() ) ;
npname.ReplaceAll("_prim","") ;
RooAbsReal* np_rfv = ws->function( npname ) ;
if ( verbLevel > 0 ) {
if ( np_rfv != 0x0 ) {
printf(" %20s : %8.2f , %15s, %8.3f\n", val->GetName(), val->getVal(), np_rfv->GetName(), np_rfv->getVal() ) ;
} else if ( strstr( npname.Data(), "eff_sf" ) != 0 ) {
np_rfv = ws->function( "eff_sf_sig" ) ;
RooAbsReal* np_rfv2 = ws->function( "eff_sf_sb" ) ;
printf(" %20s : %8.2f , %15s, %8.3f , %15s, %8.3f\n", val->GetName(), val->getVal(), np_rfv->GetName(), np_rfv->getVal(), np_rfv2->GetName(), np_rfv2->getVal() ) ;
} else if ( strstr( npname.Data(), "sf_ll" ) != 0 ) {
np_rfv = ws->function( "sf_ee" ) ;
RooAbsReal* np_rfv2 = ws->function( "sf_mm" ) ;
printf(" %20s : %8.2f , %15s, %8.3f , %15s, %8.3f\n", val->GetName(), val->getVal(), np_rfv->GetName(), np_rfv->getVal(), np_rfv2->GetName(), np_rfv2->getVal() ) ;
} else {
printf(" %20s : %8.2f\n", val->GetName(), val->getVal() ) ;
}
}
delete nprds ;
} // np_rrv iterator
}
//--- 2) Fit the dataset with these values for the nuisance parameters.
if ( isBgonlyStudy ) {
//-- fit with susy yield fixed to zero.
rrv_mu_susy_sig -> setVal( 0. ) ;
if ( verbLevel > 0 ) { printf("\n Setting mu_susy_sig to zero.\n\n") ; }
} else {
//-- fit with susy yield fixed to predicted value.
rrv_mu_susy_sig -> setVal( poiVal ) ;
if ( verbLevel > 0 ) { printf("\n Setting mu_susy_sig to %8.1f.\n\n", poiVal) ; }
}
rrv_mu_susy_sig->setConstant( kTRUE ) ;
if ( verbLevel > 0 ) {
printf("\n\n") ;
printf(" Fitting with these values for the observables to define the model for toy generation.\n") ;
rds->printMultiline(cout, 1, kTRUE, "") ;
printf("\n\n") ;
}
RooFitResult* toyModelDefinitionFitResult(0x0) ;
if ( verbLevel < 2 ) {
toyModelDefinitionFitResult = likelihood->fitTo(*rds, Save(true), PrintLevel(-1));
} else {
toyModelDefinitionFitResult = likelihood->fitTo(*rds, Save(true));
}
int toyModelDefFitCovQual = toyModelDefinitionFitResult->covQual() ;
if ( verbLevel > 0 ) { printf("\n fit covariance matrix quality: %d\n\n", toyModelDefFitCovQual ) ; }
if ( toyModelDefFitCovQual != 3 ) {
printf("\n\n\n *** Bad toy model definition fit. Cov qual %d. Aborting this toy.\n\n\n", toyModelDefFitCovQual ) ;
continue ;
}
delete toyModelDefinitionFitResult ;
if ( verbLevel > 0 ) {
printf("\n\n\n === Model values for observables. These will be used to generate the toy dataset.\n\n") ;
printObservables() ;
}
//--- 3) Generate a new set of observables based on this model.
generateObservables() ;
printf("\n\n\n Generated dataset\n") ;
rds->Print() ;
rds->printMultiline(cout, 1, kTRUE, "") ;
//--- Apparently, I need to make a new RooDataSet... Resetting the
// values in the old one doesn't stick. If you do likelihood->fitTo(*rds), it
// uses the original values, not the reset ones, in the fit.
RooArgSet toyFitobservedParametersList ;
toyFitobservedParametersList.add( *rrv_Nsig ) ;
toyFitobservedParametersList.add( *rrv_Nsb ) ;
toyFitobservedParametersList.add( *rrv_Nsig_sl ) ;
toyFitobservedParametersList.add( *rrv_Nsb_sl ) ;
toyFitobservedParametersList.add( *rrv_Nsig_ldp ) ;
toyFitobservedParametersList.add( *rrv_Nsb_ldp ) ;
toyFitobservedParametersList.add( *rrv_Nsig_ee ) ;
toyFitobservedParametersList.add( *rrv_Nsb_ee ) ;
toyFitobservedParametersList.add( *rrv_Nsig_mm ) ;
toyFitobservedParametersList.add( *rrv_Nsb_mm ) ;
RooDataSet* toyFitdsObserved = new RooDataSet("toyfit_ra2b_observed_rds", "RA2b toy observed data values",
toyFitobservedParametersList ) ;
toyFitdsObserved->add( toyFitobservedParametersList ) ;
//--- 4) Reset and free the nuisance parameters.
{
if ( verbLevel > 0 ) { printf("\n\n") ; }
int npi(0) ;
TIterator* npIter = nuisanceParameters->createIterator() ;
while ( RooRealVar* np_rrv = (RooRealVar*) npIter->Next() ) {
np_rrv -> setVal( np_initial_val[npi] ) ; // assuming that the order in the iterator does not change.
np_rrv -> setConstant( kFALSE ) ;
npi++ ;
if ( verbLevel > 0 ) { printf(" reset %20s to %8.2f and freed it.\n", np_rrv->GetName() , np_rrv->getVal() ) ; }
} // np_rrv iterator.
if ( verbLevel > 0 ) { printf("\n\n") ; }
}
//--- 5a) Evaluate the test statistic: Fit with susy yield floating to get the absolute maximum log likelihood.
if ( verbLevel > 0 ) { printf("\n\n Evaluating the test statistic for this toy. Fitting with susy floating.\n\n") ; }
rrv_mu_susy_sig->setVal( 0.0 ) ;
rrv_mu_susy_sig->setConstant( kFALSE ) ;
if ( verbLevel > 0 ) {
printf("\n toy dataset\n\n") ;
toyFitdsObserved->printMultiline(cout, 1, kTRUE, "") ;
}
/////---- nfg. Need to create a new dataset ----------
/////RooFitResult* maxLikelihoodFitResult = likelihood->fitTo(*rds, Save(true), PrintLevel(-1));
/////RooFitResult* maxLikelihoodFitResult = likelihood->fitTo(*rds, Save(true));
/////--------------
RooFitResult* maxLikelihoodFitResult(0x0) ;
if ( verbLevel < 2 ) {
maxLikelihoodFitResult = likelihood->fitTo(*toyFitdsObserved, Save(true), PrintLevel(-1));
} else {
maxLikelihoodFitResult = likelihood->fitTo(*toyFitdsObserved, Save(true));
}
if ( verbLevel > 0 ) { printObservables() ; }
int mlFitCovQual = maxLikelihoodFitResult->covQual() ;
if ( verbLevel > 0 ) { printf("\n fit covariance matrix quality: %d , -log likelihood %f\n\n", mlFitCovQual, maxLikelihoodFitResult->minNll() ) ; }
if ( mlFitCovQual != 3 ) {
printf("\n\n\n *** Bad maximum likelihood fit (susy floating). Cov qual %d. Aborting this toy.\n\n\n", mlFitCovQual ) ;
continue ;
}
double maxL_susyFloat = maxLikelihoodFitResult->minNll() ;
double maxL_mu_susy_sig = rrv_mu_susy_sig->getVal() ;
delete maxLikelihoodFitResult ;
//--- 5b) Evaluate the test statistic: Fit with susy yield fixed to hypothesis value.
// This is only necessary if the maximum likelihood fit value of the susy yield
// is less than the hypothesis value (to get a one-sided limit).
double testStat(0.0) ;
double maxL_susyFixed(0.0) ;
if ( maxL_mu_susy_sig < poiVal ) {
if ( verbLevel > 0 ) { printf("\n\n Evaluating the test statistic for this toy. Fitting with susy fixed to %8.2f.\n\n", poiVal ) ; }
rrv_mu_susy_sig->setVal( poiVal ) ;
rrv_mu_susy_sig->setConstant( kTRUE ) ;
if ( verbLevel > 0 ) {
printf("\n toy dataset\n\n") ;
rds->printMultiline(cout, 1, kTRUE, "") ;
}
////--------- nfg. need to make a new dataset ---------------
////RooFitResult* susyFixedFitResult = likelihood->fitTo(*rds, Save(true), PrintLevel(-1));
////RooFitResult* susyFixedFitResult = likelihood->fitTo(*rds, Save(true));
////-----------------------------
RooFitResult* susyFixedFitResult(0x0) ;
if ( verbLevel < 2 ) {
susyFixedFitResult = likelihood->fitTo(*toyFitdsObserved, Save(true), PrintLevel(-1));
} else {
susyFixedFitResult = likelihood->fitTo(*toyFitdsObserved, Save(true));
}
if ( verbLevel > 0 ) { printObservables() ; }
int susyFixedFitCovQual = susyFixedFitResult->covQual() ;
if ( verbLevel > 0 ) { printf("\n fit covariance matrix quality: %d , -log likelihood %f\n\n", susyFixedFitCovQual, susyFixedFitResult->minNll() ) ; }
if ( susyFixedFitCovQual != 3 ) {
printf("\n\n\n *** Bad maximum likelihood fit (susy fixed). Cov qual %d. Aborting this toy.\n\n\n", susyFixedFitCovQual ) ;
continue ;
}
maxL_susyFixed = susyFixedFitResult->minNll() ;
testStat = 2. * (maxL_susyFixed - maxL_susyFloat) ;
delete susyFixedFitResult ;
} else {
if ( verbLevel > 0 ) { printf("\n\n Floating value of susy yield greater than hypo value (%8.2f > %8.2f). Setting test stat to zero.\n\n", maxL_mu_susy_sig, poiVal ) ; }
testStat = 0.0 ;
}
printf(" --- test stat for toy %4d : %8.2f\n", ti, testStat ) ;
nToyOK++ ;
if ( testStat > dataTestStat ) { nToyWorseThanData++ ; }
if ( makeTtree ) {
tt_testStat = testStat ;
tt_gen_Nsig = rrv_Nsig->getVal() ;
tt_gen_Nsb = rrv_Nsb->getVal() ;
tt_gen_Nsig_sl = rrv_Nsig_sl->getVal() ;
tt_gen_Nsb_sl = rrv_Nsb_sl->getVal() ;
tt_gen_Nsig_ldp = rrv_Nsig_ldp->getVal() ;
tt_gen_Nsb_ldp = rrv_Nsb_ldp->getVal() ;
tt_gen_Nsig_ee = rrv_Nsig_ee->getVal() ;
tt_gen_Nsb_ee = rrv_Nsb_ee->getVal() ;
tt_gen_Nsig_mm = rrv_Nsig_mm->getVal() ;
tt_gen_Nsb_mm = rrv_Nsb_mm->getVal() ;
toytt->Fill() ;
}
//--- *) reset things for the next toy.
resetObservables() ;
delete toyFitdsObserved ;
} // ti.
wstf->Close() ;
printf("\n\n\n") ;
if ( nToyOK == 0 ) { printf("\n\n\n *** All toys bad !?!?!\n\n\n") ; return ; }
double pValue = (1.0*nToyWorseThanData) / (1.0*nToyOK) ;
if ( isBgonlyStudy ) {
printf("\n\n\n p-value result, BG-only , poi=%3.0f : %4d / %4d = %6.3f\n\n\n\n", poiVal, nToyWorseThanData, nToyOK, pValue ) ;
} else {
printf("\n\n\n p-value result, S-plus-B, poi=%3.0f : %4d / %4d = %6.3f\n\n\n\n", poiVal, nToyWorseThanData, nToyOK, pValue ) ;
}
if ( makeTtree ) {
printf("\n\n Writing TTree : %s : %s\n\n", ttname, tttitle ) ;
ttfile->cd() ;
toytt->Write() ;
}
} // ws_cls_hybrid1
