| 1 |
#include "TROOT.h"
|
| 2 |
#include "TH1D.h"
|
| 3 |
#include "RooDataSet.h"
|
| 4 |
#include "RooRealVar.h"
|
| 5 |
#include "RooAbsPdf.h"
|
| 6 |
#include "RooBreitWigner.h"
|
| 7 |
#include "RooCBShape.h"
|
| 8 |
#include "RooGaussian.h"
|
| 9 |
#include "RooFFTConvPdf.h"
|
| 10 |
#include "RooDataHist.h"
|
| 11 |
#include "RooHistPdf.h"
|
| 12 |
#include "RooVoigtianShape.h"
|
| 13 |
#include "RooKeysPdf.h"
|
| 14 |
|
| 15 |
class CSignalModel
|
| 16 |
{
|
| 17 |
public:
|
| 18 |
CSignalModel():model(0){}
|
| 19 |
virtual ~CSignalModel(){ delete model; }
|
| 20 |
RooAbsPdf *model;
|
| 21 |
};
|
| 22 |
|
| 23 |
class CBreitWignerConvCrystalBall : public CSignalModel
|
| 24 |
{
|
| 25 |
public:
|
| 26 |
CBreitWignerConvCrystalBall(RooRealVar &m, const Bool_t pass);
|
| 27 |
~CBreitWignerConvCrystalBall();
|
| 28 |
RooRealVar *mass, *width;
|
| 29 |
RooBreitWigner *bw;
|
| 30 |
RooRealVar *mean, *sigma, *alpha, *n;
|
| 31 |
RooCBShape *cb;
|
| 32 |
};
|
| 33 |
|
| 34 |
class CMCTemplateConvGaussian : public CSignalModel
|
| 35 |
{
|
| 36 |
public:
|
| 37 |
CMCTemplateConvGaussian(RooRealVar &m, TH1D* hist, const Bool_t pass, RooRealVar *sigma0=0, int intOrder=1);
|
| 38 |
~CMCTemplateConvGaussian();
|
| 39 |
RooRealVar *mean, *sigma;
|
| 40 |
RooGaussian *gaus;
|
| 41 |
TH1D *inHist;
|
| 42 |
RooDataHist *dataHist;
|
| 43 |
RooHistPdf *histPdf;
|
| 44 |
};
|
| 45 |
|
| 46 |
class CVoigtianCBShape : public CSignalModel
|
| 47 |
{
|
| 48 |
public:
|
| 49 |
CVoigtianCBShape(RooRealVar &m, const Bool_t pass);
|
| 50 |
~CVoigtianCBShape();
|
| 51 |
RooRealVar *mass, *width;
|
| 52 |
RooRealVar *sigma, *alpha, *n;
|
| 53 |
};
|
| 54 |
|
| 55 |
class CMCDatasetConvGaussian : public CSignalModel
|
| 56 |
{
|
| 57 |
public:
|
| 58 |
CMCDatasetConvGaussian(RooRealVar &m, TTree* tree, const Bool_t pass, RooRealVar *sigma0=0);
|
| 59 |
~CMCDatasetConvGaussian();
|
| 60 |
RooRealVar *mean, *sigma;
|
| 61 |
RooGaussian *gaus;
|
| 62 |
TTree *inTree;
|
| 63 |
RooDataSet *dataSet;
|
| 64 |
RooKeysPdf *keysPdf;
|
| 65 |
};
|
| 66 |
|
| 67 |
//--------------------------------------------------------------------------------------------------
|
| 68 |
CBreitWignerConvCrystalBall::CBreitWignerConvCrystalBall(RooRealVar &m, const Bool_t pass)
|
| 69 |
{
|
| 70 |
char name[10];
|
| 71 |
if(pass) sprintf(name,"%s","Pass");
|
| 72 |
else sprintf(name,"%s","Fail");
|
| 73 |
|
| 74 |
char vname[50];
|
| 75 |
|
| 76 |
sprintf(vname,"mass%s",name);
|
| 77 |
mass = new RooRealVar(vname,vname,91,80,100);
|
| 78 |
mass->setVal(91.1876);
|
| 79 |
mass->setConstant(kTRUE);
|
| 80 |
|
| 81 |
sprintf(vname,"width%s",name);
|
| 82 |
width = new RooRealVar(vname,vname,2.5,0.1,10);
|
| 83 |
width->setVal(2.4952);
|
| 84 |
width->setConstant(kTRUE);
|
| 85 |
|
| 86 |
sprintf(vname,"bw%s",name);
|
| 87 |
bw = new RooBreitWigner(vname,vname,m,*mass,*width);
|
| 88 |
|
| 89 |
if(pass) {
|
| 90 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 91 |
sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,1,0.1,5);
|
| 92 |
sprintf(vname,"alpha%s",name); alpha = new RooRealVar(vname,vname,5,0,20);
|
| 93 |
sprintf(vname,"n%s",name); n = new RooRealVar(vname,vname,1,0,10);
|
| 94 |
} else {
|
| 95 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 96 |
sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,1,0.1,5);
|
| 97 |
sprintf(vname,"alpha%s",name); alpha = new RooRealVar(vname,vname,5,0,20);
|
| 98 |
sprintf(vname,"n%s",name); n = new RooRealVar(vname,vname,1,0,10);
|
| 99 |
}
|
| 100 |
// n->setVal(1.0);
|
| 101 |
// n->setConstant(kTRUE);
|
| 102 |
|
| 103 |
sprintf(vname,"cb%s",name);
|
| 104 |
cb = new RooCBShape(vname,vname,m,*mean,*sigma,*alpha,*n);
|
| 105 |
|
| 106 |
sprintf(vname,"signal%s",name);
|
| 107 |
model = new RooFFTConvPdf(vname,vname,m,*bw,*cb);
|
| 108 |
}
|
| 109 |
|
| 110 |
CBreitWignerConvCrystalBall::~CBreitWignerConvCrystalBall()
|
| 111 |
{
|
| 112 |
delete mass;
|
| 113 |
delete width;
|
| 114 |
delete bw;
|
| 115 |
delete mean;
|
| 116 |
delete sigma;
|
| 117 |
delete alpha;
|
| 118 |
delete n;
|
| 119 |
delete cb;
|
| 120 |
}
|
| 121 |
|
| 122 |
//--------------------------------------------------------------------------------------------------
|
| 123 |
CMCTemplateConvGaussian::CMCTemplateConvGaussian(RooRealVar &m, TH1D* hist, const Bool_t pass, RooRealVar *sigma0, int intOrder)
|
| 124 |
{
|
| 125 |
char name[10];
|
| 126 |
if(pass) sprintf(name,"%s","Pass");
|
| 127 |
else sprintf(name,"%s","Fail");
|
| 128 |
|
| 129 |
char vname[50];
|
| 130 |
|
| 131 |
if(pass) {
|
| 132 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 133 |
if(sigma0) { sigma = sigma0; }
|
| 134 |
else { sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,2,0,5); }
|
| 135 |
sprintf(vname,"gaus%s",name); gaus = new RooGaussian(vname,vname,m,*mean,*sigma);
|
| 136 |
} else {
|
| 137 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 138 |
if(sigma0) { sigma = sigma0; }
|
| 139 |
else { sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,2,0,5); }
|
| 140 |
sprintf(vname,"gaus%s",name); gaus = new RooGaussian(vname,vname,m,*mean,*sigma);
|
| 141 |
}
|
| 142 |
|
| 143 |
|
| 144 |
sprintf(vname,"inHist_%s",hist->GetName());
|
| 145 |
inHist = (TH1D*)hist->Clone(vname);
|
| 146 |
|
| 147 |
sprintf(vname,"dataHist%s",name); dataHist = new RooDataHist(vname,vname,RooArgSet(m),inHist);
|
| 148 |
sprintf(vname,"histPdf%s",name); histPdf = new RooHistPdf(vname,vname,m,*dataHist,intOrder);
|
| 149 |
sprintf(vname,"signal%s",name); model = new RooFFTConvPdf(vname,vname,m,*histPdf,*gaus);
|
| 150 |
}
|
| 151 |
|
| 152 |
CMCTemplateConvGaussian::~CMCTemplateConvGaussian()
|
| 153 |
{
|
| 154 |
delete mean;
|
| 155 |
//delete sigma;
|
| 156 |
delete gaus;
|
| 157 |
delete inHist;
|
| 158 |
delete dataHist;
|
| 159 |
delete histPdf;
|
| 160 |
}
|
| 161 |
|
| 162 |
//--------------------------------------------------------------------------------------------------
|
| 163 |
CVoigtianCBShape::CVoigtianCBShape(RooRealVar &m, const Bool_t pass)
|
| 164 |
{
|
| 165 |
char name[10];
|
| 166 |
if(pass) sprintf(name,"%s","Pass");
|
| 167 |
else sprintf(name,"%s","Fail");
|
| 168 |
|
| 169 |
char vname[50];
|
| 170 |
|
| 171 |
sprintf(vname,"mass%s",name);
|
| 172 |
mass = new RooRealVar(vname,vname,91,80,100);
|
| 173 |
|
| 174 |
sprintf(vname,"width%s",name);
|
| 175 |
width = new RooRealVar(vname,vname,2.5,0.1,10);
|
| 176 |
width->setVal(2.4952);
|
| 177 |
width->setConstant(kTRUE);
|
| 178 |
|
| 179 |
if(pass) {
|
| 180 |
sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,1,0.1,10);
|
| 181 |
sprintf(vname,"alpha%s",name); alpha = new RooRealVar(vname,vname,5,0,20);
|
| 182 |
} else {
|
| 183 |
sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,1,0.1,10);
|
| 184 |
sprintf(vname,"alpha%s",name); alpha = new RooRealVar(vname,vname,5,0,20);
|
| 185 |
}
|
| 186 |
|
| 187 |
sprintf(vname,"n%s",name);
|
| 188 |
n = new RooRealVar(vname,vname,1,0,10);
|
| 189 |
n->setVal(1.0);
|
| 190 |
|
| 191 |
sprintf(vname,"signal%s",name);
|
| 192 |
model = new RooVoigtianShape(vname,vname,m,*mass,*sigma,*alpha,*n,*width,0);
|
| 193 |
}
|
| 194 |
|
| 195 |
CVoigtianCBShape::~CVoigtianCBShape()
|
| 196 |
{
|
| 197 |
delete mass;
|
| 198 |
delete width;
|
| 199 |
delete sigma;
|
| 200 |
delete alpha;
|
| 201 |
delete n;
|
| 202 |
}
|
| 203 |
|
| 204 |
//--------------------------------------------------------------------------------------------------
|
| 205 |
CMCDatasetConvGaussian::CMCDatasetConvGaussian(RooRealVar &m, TTree* tree, const Bool_t pass, RooRealVar *sigma0)
|
| 206 |
{
|
| 207 |
char name[10];
|
| 208 |
if(pass) sprintf(name,"%s","Pass");
|
| 209 |
else sprintf(name,"%s","Fail");
|
| 210 |
|
| 211 |
char vname[50];
|
| 212 |
|
| 213 |
if(pass) {
|
| 214 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 215 |
if(sigma0) { sigma = sigma0; }
|
| 216 |
else { sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,2,0,5); }
|
| 217 |
sprintf(vname,"gaus%s",name); gaus = new RooGaussian(vname,vname,m,*mean,*sigma);
|
| 218 |
} else {
|
| 219 |
sprintf(vname,"mean%s",name); mean = new RooRealVar(vname,vname,0,-10,10);
|
| 220 |
if(sigma0) { sigma = sigma0; }
|
| 221 |
else { sprintf(vname,"sigma%s",name); sigma = new RooRealVar(vname,vname,2,0,5); }
|
| 222 |
sprintf(vname,"gaus%s",name); gaus = new RooGaussian(vname,vname,m,*mean,*sigma);
|
| 223 |
}
|
| 224 |
|
| 225 |
sprintf(vname,"inTree_%s",tree->GetName());
|
| 226 |
inTree = (TTree*)tree->Clone(vname);
|
| 227 |
|
| 228 |
sprintf(vname,"dataSet%s",name); dataSet = new RooDataSet(vname,vname,inTree,RooArgSet(m));
|
| 229 |
sprintf(vname,"keysPdf%s",name); keysPdf = new RooKeysPdf(vname,vname,m,*dataSet,RooKeysPdf::NoMirror,1);
|
| 230 |
sprintf(vname,"signal%s",name); model = new RooFFTConvPdf(vname,vname,m,*keysPdf,*gaus);
|
| 231 |
}
|
| 232 |
|
| 233 |
CMCDatasetConvGaussian::~CMCDatasetConvGaussian()
|
| 234 |
{
|
| 235 |
delete mean;
|
| 236 |
delete sigma;
|
| 237 |
delete gaus;
|
| 238 |
delete inTree;
|
| 239 |
delete dataSet;
|
| 240 |
delete keysPdf;
|
| 241 |
}
|
