CMS 3D CMS Logo

MuonResidualsBfieldAngleFitter.cc
Go to the documentation of this file.
2 
4 
5 void MuonResidualsBfieldAngleFitter::inform(TMinuit *tMinuit) {
7 }
8 
9 void MuonResidualsBfieldAngleFitter_FCN(int &npar, double *gin, double &fval, double *par, int iflag) {
11  MuonResidualsFitter *fitter = fitinfo->fitter();
12 
13  fval = 0.;
14  for (std::vector<double*>::const_iterator resiter = fitter->residuals_begin(); resiter != fitter->residuals_end(); ++resiter) {
15  const double residual = (*resiter)[MuonResidualsBfieldAngleFitter::kResidual];
16  const double qoverpt = (*resiter)[MuonResidualsBfieldAngleFitter::kQoverPt];
17  const double qoverpz = (*resiter)[MuonResidualsBfieldAngleFitter::kQoverPz];
18 
19  double center = 0.;
21  center += par[MuonResidualsBfieldAngleFitter::kBfrompt] * qoverpt;
22  center += par[MuonResidualsBfieldAngleFitter::kBfrompz] * qoverpz;
23  center += par[MuonResidualsBfieldAngleFitter::kdEdx] * (1./qoverpt/qoverpt + 1./qoverpz/qoverpz) * (qoverpt > 0. ? 1. : -1.);
24 
27  }
30  }
31  else if (fitter->residualsModel() == MuonResidualsFitter::kROOTVoigt) {
33  }
36  }
37  else { assert(false); }
38  }
39 }
40 
42  initialize_table(); // if not already initialized
43 
44  double sum_x = 0.;
45  double sum_xx = 0.;
46  int N = 0;
47 
48  for (std::vector<double*>::const_iterator resiter = residuals_begin(); resiter != residuals_end(); ++resiter) {
49  const double residual = (*resiter)[kResidual];
50 // const double qoverpt = (*resiter)[kQoverPt];
51 
52  if (fabs(residual) < 0.1) { // truncate at 100 mrad
53  sum_x += residual;
54  sum_xx += residual*residual;
55  N++;
56  }
57  }
58 
59  if (N < m_minHits) return false;
60 
61  // truncated mean and stdev to seed the fit
62  double mean = sum_x/double(N);
63  double stdev = sqrt(sum_xx/double(N) - pow(sum_x/double(N), 2));
64 
65  // refine the standard deviation calculation
66  sum_x = 0.;
67  sum_xx = 0.;
68  N = 0;
69  for (std::vector<double*>::const_iterator resiter = residuals_begin(); resiter != residuals_end(); ++resiter) {
70  const double residual = (*resiter)[kResidual];
71  if (mean - 1.5*stdev < residual && residual < mean + 1.5*stdev) {
72  sum_x += residual;
73  sum_xx += residual*residual;
74  N++;
75  }
76  }
77  mean = sum_x/double(N);
78  stdev = sqrt(sum_xx/double(N) - pow(sum_x/double(N), 2));
79 
80  sum_x = 0.;
81  sum_xx = 0.;
82  N = 0;
83  for (std::vector<double*>::const_iterator resiter = residuals_begin(); resiter != residuals_end(); ++resiter) {
84  const double residual = (*resiter)[kResidual];
85  if (mean - 1.5*stdev < residual && residual < mean + 1.5*stdev) {
86  sum_x += residual;
87  sum_xx += residual*residual;
88  N++;
89  }
90  }
91  mean = sum_x/double(N);
92  stdev = sqrt(sum_xx/double(N) - pow(sum_x/double(N), 2));
93 
94  std::vector<int> parNum;
95  std::vector<std::string> parName;
96  std::vector<double> start;
97  std::vector<double> step;
98  std::vector<double> low;
99  std::vector<double> high;
100 
101  parNum.push_back(kAngle); parName.push_back(std::string("angle")); start.push_back(mean); step.push_back(0.1); low.push_back(0.); high.push_back(0.);
102  parNum.push_back(kBfrompt); parName.push_back(std::string("bfrompt")); start.push_back(0.); step.push_back(0.1*stdev/0.05); low.push_back(0.); high.push_back(0.);
103  parNum.push_back(kBfrompz); parName.push_back(std::string("bfrompz")); start.push_back(0.); step.push_back(0.1*stdev/0.05); low.push_back(0.); high.push_back(0.);
104  parNum.push_back(kdEdx); parName.push_back(std::string("dEdx")); start.push_back(0.); step.push_back(0.1*stdev/0.05); low.push_back(0.); high.push_back(0.);
105  parNum.push_back(kSigma); parName.push_back(std::string("sigma")); start.push_back(stdev); step.push_back(0.1*stdev); low.push_back(0.); high.push_back(0.);
107  parNum.push_back(kGamma); parName.push_back(std::string("gamma")); start.push_back(stdev); step.push_back(0.1*stdev); low.push_back(0.); high.push_back(0.);
108  }
109 
110  return dofit(&MuonResidualsBfieldAngleFitter_FCN, parNum, parName, start, step, low, high);
111 }
112 
114  std::stringstream raw_name, narrowed_name, qoverpt_name, qoverpz_name, psquared_name;
115  raw_name << name << "_raw";
116  narrowed_name << name << "_narrowed";
117  qoverpt_name << name << "_qoverpt";
118  qoverpz_name << name << "_qoverpz";
119  psquared_name << name << "_psquared";
120 
121  TH1F *raw_hist = dir->make<TH1F>(raw_name.str().c_str(), (raw_name.str() + std::string(" (mrad)")).c_str(), 100, -100., 100.);
122  TH1F *narrowed_hist = dir->make<TH1F>(narrowed_name.str().c_str(), (narrowed_name.str() + std::string(" (mrad)")).c_str(), 100, -100., 100.);
123  TProfile *qoverpt_hist = dir->make<TProfile>(qoverpt_name.str().c_str(), (qoverpt_name.str() + std::string(" (mrad)")).c_str(), 100, -0.05, 0.05);
124  TProfile *qoverpz_hist = dir->make<TProfile>(qoverpz_name.str().c_str(), (qoverpz_name.str() + std::string(" (mrad)")).c_str(), 100, -0.05, 0.05);
125  TProfile *psquared_hist = dir->make<TProfile>(psquared_name.str().c_str(), (psquared_name.str() + std::string(" (mrad)")).c_str(), 100, -0.05, 0.05);
126 
127  narrowed_name << "fit";
128  qoverpt_name << "fit";
129  qoverpz_name << "fit";
130  psquared_name << "fit";
131 
132  double scale_factor = double(numResiduals()) * (100. - -100.)/100; // (max - min)/nbins
133 
134  TF1 *narrowed_fit = nullptr;
135  if (residualsModel() == kPureGaussian) {
136  narrowed_fit = new TF1(narrowed_name.str().c_str(), MuonResidualsFitter_pureGaussian_TF1, -100., 100., 3);
137  narrowed_fit->SetParameters(scale_factor, value(kAngle) * 1000., value(kSigma) * 1000.);
138  narrowed_fit->Write();
139  }
140  else if (residualsModel() == kPowerLawTails) {
141  narrowed_fit = new TF1(narrowed_name.str().c_str(), MuonResidualsFitter_powerLawTails_TF1, -100., 100., 4);
142  narrowed_fit->SetParameters(scale_factor, value(kAngle) * 1000., value(kSigma) * 1000., value(kGamma) * 1000.);
143  narrowed_fit->Write();
144  }
145  else if (residualsModel() == kROOTVoigt) {
146  narrowed_fit = new TF1(narrowed_name.str().c_str(), MuonResidualsFitter_ROOTVoigt_TF1, -100., 100., 4);
147  narrowed_fit->SetParameters(scale_factor, value(kAngle) * 1000., value(kSigma) * 1000., value(kGamma) * 1000.);
148  narrowed_fit->Write();
149  }
150  else if (residualsModel() == kGaussPowerTails) {
151  narrowed_fit = new TF1(narrowed_name.str().c_str(), MuonResidualsFitter_GaussPowerTails_TF1, -100., 100., 3);
152  narrowed_fit->SetParameters(scale_factor, value(kAngle) * 1000., value(kSigma) * 1000.);
153  narrowed_fit->Write();
154  }
155 
156  TF1 *qoverpt_fit = new TF1(qoverpt_name.str().c_str(), "[0]+x*[1]", -0.05, 0.05);
157  qoverpt_fit->SetParameters(value(kAngle) * 1000., value(kBfrompt) * 1000.);
158  qoverpt_fit->Write();
159 
160  TF1 *qoverpz_fit = new TF1(qoverpz_name.str().c_str(), "[0]+x*[1]", -0.05, 0.05);
161  qoverpz_fit->SetParameters(value(kAngle) * 1000., value(kBfrompz) * 1000.);
162  qoverpz_fit->Write();
163 
164  TF1 *psquared_fit = new TF1(psquared_name.str().c_str(), "[0]+[1]*x**2", -0.05, 0.05);
165  psquared_fit->SetParameters(value(kAngle) * 1000., value(kdEdx) * 1000.);
166  psquared_fit->Write();
167 
168  for (std::vector<double*>::const_iterator resiter = residuals_begin(); resiter != residuals_end(); ++resiter) {
169  const double raw_residual = (*resiter)[kResidual];
170  const double qoverpt = (*resiter)[kQoverPt];
171  const double qoverpz = (*resiter)[kQoverPz];
172  const double psquared = (1./qoverpt/qoverpt + 1./qoverpz/qoverpz) * (qoverpt > 0. ? 1. : -1.);
173 
174  double qoverpt_correction = value(kBfrompt) * qoverpt;
175  double qoverpz_correction = value(kBfrompz) * qoverpz;
176  double dEdx_correction = value(kdEdx) * psquared;
177  double corrected_residual = raw_residual - qoverpt_correction - qoverpz_correction - dEdx_correction;
178 
179  raw_hist->Fill(raw_residual * 1000.);
180  narrowed_hist->Fill(corrected_residual * 1000.);
181 
182  qoverpt_hist->Fill(qoverpt, (raw_residual - qoverpz_correction - dEdx_correction) * 1000.);
183  qoverpz_hist->Fill(qoverpz, (raw_residual - qoverpt_correction - dEdx_correction) * 1000.);
184  psquared_hist->Fill(psquared, (raw_residual - qoverpt_correction - qoverpz_correction) * 1000.);
185  }
186 
187  return 0.;
188 }
double MuonResidualsFitter_logGaussPowerTails(double residual, double center, double sigma)
Double_t MuonResidualsFitter_powerLawTails_TF1(Double_t *xvec, Double_t *par)
MuonResidualsFitter * fitter()
double value(int parNum)
bool dofit(void(*fcn)(int &, double *, double &, double *, int), std::vector< int > &parNum, std::vector< std::string > &parName, std::vector< double > &start, std::vector< double > &step, std::vector< double > &low, std::vector< double > &high)
double MuonResidualsFitter_logROOTVoigt(double residual, double center, double sigma, double gamma)
void inform(TMinuit *tMinuit) override
double plot(std::string name, TFileDirectory *dir, Alignable *ali) override
T sqrt(T t)
Definition: SSEVec.h:18
double MuonResidualsFitter_logPowerLawTails(double residual, double center, double sigma, double gamma)
Double_t MuonResidualsFitter_GaussPowerTails_TF1(Double_t *xvec, Double_t *par)
T * make(const Args &...args) const
make new ROOT object
Double_t MuonResidualsFitter_pureGaussian_TF1(Double_t *xvec, Double_t *par)
#define N
Definition: blowfish.cc:9
static TMinuit * MuonResidualsBfieldAngleFitter_TMinuit
std::vector< double * >::const_iterator residuals_end() const
def stdev(xlist)
Definition: plotscripts.py:69
step
Definition: StallMonitor.cc:94
dbl *** dir
Definition: mlp_gen.cc:35
Double_t MuonResidualsFitter_ROOTVoigt_TF1(Double_t *xvec, Double_t *par)
std::vector< double * >::const_iterator residuals_begin() const
void MuonResidualsBfieldAngleFitter_FCN(int &npar, double *gin, double &fval, double *par, int iflag)
double MuonResidualsFitter_logPureGaussian(double residual, double center, double sigma)
Power< A, B >::type pow(const A &a, const B &b)
Definition: Power.h:40