MomentumScaleCorrector Class Reference

#include <MomentumScaleCorrector.h>

Inheritance diagram for MomentumScaleCorrector:

Public Member Functions
template<class U >
double	correct (const U &lorentzVector)
	Alternative method that can be used with lorentzVectors. More...
scaleFunctionBase< double * > *	function (const int i)
	Returns a pointer to the selected function. More...
	MomentumScaleCorrector (TString identifier)
	MomentumScaleCorrector (const MuScleFitDBobject *dbObject)
template<class U >
double	operator() (const U &track)
	Method to do the corrections. It is templated to work with all the track types. More...
	~MomentumScaleCorrector ()
Public Member Functions inherited from BaseFunction
	BaseFunction ()
	BaseFunction (const MuScleFitDBobject *dbObject)
	Constructor when receiving database parameters. More...
std::vector< double >	fitQuality () const
	Return the vector of fit quality values. More...
std::vector< int >	identifiers () const
	Return the vector of function identifiers. More...
std::vector< double >	parameters () const
	Return the vector of parameters. More...

Protected Member Functions
void	readParameters (TString fileName)
	Parser of the parameters file. More...
Protected Member Functions inherited from BaseFunction
template<class T >
void	convertToArrays (T **&function_, const std::vector< T * > &functionVec_)
	Convert vectors to arrays for faster random access. The first pointer is replaced, thus it is taken by reference. More...

Protected Attributes
scaleFunctionBase< double * > **	scaleFunction_
std::vector< scaleFunctionBase< double * > * >	scaleFunctionVec_
Protected Attributes inherited from BaseFunction
std::vector< double >	fitQuality_
double **	fitQualityArray_
std::vector< int >	functionId_
int	iterationNum_
double **	parArray_
std::vector< double >	parVecVec_

Detailed Description

MomentumScaleCorrector class Author M. De Mattia - 18/11/2008 This is used to have a common set of functions for the specialized templates to use. The constructor receives the name identifying the parameters for the correction function. It reads the parameters from a txt file in data/.

It handles multiple iterations. It is also possible to use different functions in different iterations.

ATTENTION: it is important that iterations numbers in the txt file start from 0.

Definition at line 25 of file MomentumScaleCorrector.h.

Constructor & Destructor Documentation

MomentumScaleCorrector::MomentumScaleCorrector ( TString  identifier )

inline

The constructor takes a string identifying the parameters to read. It parses the txt file containing the parameters, extracts the index of the correction function and saves the corresponding pointer. It then fills the vector of parameters.

Definition at line 33 of file MomentumScaleCorrector.h.

References readParameters().

                                             {
    identifier.Prepend("MuonAnalysis/MomentumScaleCalibration/data/");
    identifier.Append(".txt");
    edm::FileInPath fileWithFullPath(identifier.Data());
    readParameters(fileWithFullPath.fullPath());
  }

MomentumScaleCorrector::MomentumScaleCorrector ( const MuScleFitDBobject *  dbObject )

inline

This constructor is used when reading parameters from the db. It receives a pointer to an object of type MuScleFitDBobject containing the parameters and the functions identifiers.

Definition at line 44 of file MomentumScaleCorrector.h.

References BaseFunction::convertToArrays(), BaseFunction::functionId_, triggerObjects_cff::id, scaleFunction_, scaleFunctionService(), and scaleFunctionVec_.

                                                            : BaseFunction(dbObject) {
    std::vector<int>::const_iterator id = functionId_.begin();
    for (; id != functionId_.end(); ++id) {
      scaleFunctionVec_.push_back(scaleFunctionService(*id));
    }
    // Fill the arrays that will be used when calling the correction function.
    convertToArrays(scaleFunction_, scaleFunctionVec_);
  }

MomentumScaleCorrector::~MomentumScaleCorrector ( )

inline

Definition at line 53 of file MomentumScaleCorrector.h.

References BaseFunction::functionId_, mps_fire::i, BaseFunction::parArray_, and scaleFunction_.

                            {
    if (parArray_ != nullptr) {
      for (unsigned int i = 0; i < functionId_.size(); ++i) {
        delete[] parArray_[i];
        delete scaleFunction_[i];
      }
      delete[] parArray_;
      delete[] scaleFunction_;
    }
  }

Member Function Documentation

template<class U >

double MomentumScaleCorrector::correct ( const U &  lorentzVector )

inline

Alternative method that can be used with lorentzVectors.

Definition at line 81 of file MomentumScaleCorrector.h.

References MillePedeFileConverter_cfg::fileName, mps_fire::i, BaseFunction::iterationNum_, BaseFunction::parArray_, DiDispStaMuonMonitor_cfi::pt, readParameters(), scaleFunctionBase< T >::scale(), and scaleFunction_.

Referenced by KalmanMuonCorrector.KalmanMuonCorrector::correct_all(), and JetReCalibrator.JetReCalibrator::correctAll().

                                         {
    // Loop on all the functions and apply them iteratively on the pt corrected by the previous function.
    double pt = lorentzVector.Pt();
    for (int i = 0; i <= iterationNum_; ++i) {
      pt = (scaleFunction_[i]->scale(pt, lorentzVector.Eta(), lorentzVector.Phi(), 1, parArray_[i]));
    }
    return pt;
  }

scaleFunctionBase<double*>* MomentumScaleCorrector::function ( const int  i )

inline

Returns a pointer to the selected function.

Definition at line 65 of file MomentumScaleCorrector.h.

References mps_fire::i, and scaleFunction_.

{ return scaleFunction_[i]; }

template<class U >

double MomentumScaleCorrector::operator() ( const U &  track )

inline

Method to do the corrections. It is templated to work with all the track types.

Definition at line 69 of file MomentumScaleCorrector.h.

References mps_fire::i, BaseFunction::iterationNum_, BaseFunction::parArray_, DiDispStaMuonMonitor_cfi::pt, scaleFunctionBase< T >::scale(), and scaleFunction_.

                                    {
    // Loop on all the functions and apply them iteratively on the pt corrected by the previous function.
    double pt = track.pt();
    for (int i = 0; i <= iterationNum_; ++i) {
      // return ( scaleFunction_->scale( track.pt(), track.eta(), track.phi(), track.charge(), parScale_) );
      pt = (scaleFunction_[i]->scale(pt, track.eta(), track.phi(), track.charge(), parArray_[i]));
    }
    return pt;
  }

void MomentumScaleCorrector::readParameters ( TString  fileName )

protected

Parser of the parameters file.

Definition at line 3 of file MomentumScaleCorrector.cc.

References BaseFunction::convertToArrays(), gather_cfg::cout, BaseFunction::functionId_, recoMuon::in, align_cfg::iteration, BaseFunction::iterationNum_, mps_splice::line, EgammaValidation_cff::num, BaseFunction::parArray_, BaseFunction::parVecVec_, scaleFunction_, scaleFunctionService(), scaleFunctionVec_, and AlCaHLTBitMon_QueryRunRegistry::string.

Referenced by correct(), and MomentumScaleCorrector().

                                                            {
  iterationNum_ = 0;
  parArray_ = nullptr;
  // std::vector<double> parameterErrors;

  // Read the parameters file
  std::ifstream parametersFile(fileName.Data());

  if (!parametersFile.is_open()) {
    std::cout << "Error: file " << fileName << " not found. Aborting." << std::endl;
    abort();
  }
  std::string line;

  std::string iteration("Iteration ");
  // Loop on the file lines
  while (parametersFile) {
    getline(parametersFile, line);
    size_t lineInt = line.find("value");

    // if( line.find(iteration) != std::string::npos ) {
    size_t iterationSubStr = line.find(iteration);

    // Take the iteration number
    if (iterationSubStr != std::string::npos) {
      int scaleFunctionNum = 0;
      // This can be used when dealing with multiple iterations

      // std::cout << "line = " << line << std::endl;
      std::stringstream sLine(line);
      std::string num;
      int wordCounter = 0;
      // Warning: this strongly depends on the parameters file structure.
      while (sLine >> num) {
        ++wordCounter;
        //         std::cout << "num["<<wordCounter<<"] = " << num << std::endl;
        if (wordCounter == 9) {
          std::stringstream in(num);
          in >> scaleFunctionNum;
        }
        if (wordCounter == 13) {
          std::stringstream in(num);
          in >> iterationNum_;
        }
      }
      // std::cout << "iteration number = " << iterationNum_ << std::endl;
      // std::cout << "scale function number = " << scaleFunctionNum << std::endl;

      // Create a new vector to hold the parameters for this iteration
      //       std::vector<double> parScale;
      //       parVecVec_.push_back(parScale);

      // Set the scaleFunction
      // scaleFunction_ = scaleFunctionArrayForVec[scaleFunctionNum];
      // scaleFunction_ = scaleFunctionArray[scaleFunctionNum];
      functionId_.push_back(scaleFunctionNum);
      // scaleFunctionVec_.push_back( scaleFunctionArray[scaleFunctionNum] );
      scaleFunctionVec_.push_back(scaleFunctionService(scaleFunctionNum));
    }
    // Take the parameters for the current iteration
    if ((lineInt != std::string::npos)) {
      size_t subStr1 = line.find("value");
      std::stringstream paramStr;
      double param = 0;
      // Even if all the rest of the line is taken, the following
      // conversion to a double will stop at the end of the first number.
      paramStr << line.substr(subStr1 + 5);
      paramStr >> param;
      //       // Fill the last vector of parameters, which corresponds to this iteration.
      //       parVecVec_.back().push_back(param);
      parVecVec_.push_back(param);
      // std::cout << "param = " << param << std::endl;

      // This is to extract parameter errors
      // size_t subStr2 = line.find("+-");
      // std::stringstream parErrorStr;
      // double parError = 0;
      // parErrorStr << line.substr(subStr2+1);
      // parErrorStr >> parError;
      // parameterErrors.push_back(parError);
      // std::cout << "parError = " << parError << std::endl;
    }
  }

  convertToArrays(scaleFunction_, scaleFunctionVec_);
}

Member Data Documentation

scaleFunctionBase<double*>** MomentumScaleCorrector::scaleFunction_

protected

Definition at line 94 of file MomentumScaleCorrector.h.

Referenced by correct(), function(), MomentumScaleCorrector(), operator()(), readParameters(), and ~MomentumScaleCorrector().

std::vector<scaleFunctionBase<double*>*> MomentumScaleCorrector::scaleFunctionVec_

protected

Definition at line 95 of file MomentumScaleCorrector.h.

Referenced by MomentumScaleCorrector(), and readParameters().