#include <EcalUncalibRecHitRecChi2Algo.h>

Public Member Functions
virtual double	chi2 ()

virtual double	chi2OutOfTime ()

	EcalUncalibRecHitRecChi2Algo ()

	EcalUncalibRecHitRecChi2Algo (const C &dataFrame, const double amplitude, const EcalTimeCalibConstant &timeIC, const double amplitudeOutOfTime, const double jitter, const double pedestals, const double pedestalsRMS, const double *gainRatios, const EcalShapeBase &testbeamPulseShape, const std::vector< double > &chi2Parameters)

virtual	~EcalUncalibRecHitRecChi2Algo ()

Private Attributes
double	chi2_

double	chi2OutOfTime_

Detailed Description

template<class C>
class EcalUncalibRecHitRecChi2Algo< C >

Template used to compute the chi2 of an MGPA pulse for in-time and out-of-time signals, algorithm based on the chi2express. The in-time chi2 is calculated against the time intercalibrations from the DB while the out-of-time chi2 is calculated against the Tmax measurement on event by event basis.

Author: Konstantinos Theofilatos 02 Feb 2010

Definition at line 24 of file EcalUncalibRecHitRecChi2Algo.h.

Constructor & Destructor Documentation

template<class C>

virtual EcalUncalibRecHitRecChi2Algo< C >::~EcalUncalibRecHitRecChi2Algo ( )

inlinevirtual

Definition at line 29 of file EcalUncalibRecHitRecChi2Algo.h.

29 { };

template<class C>

EcalUncalibRecHitRecChi2Algo< C >::EcalUncalibRecHitRecChi2Algo ( )

inline

Definition at line 31 of file EcalUncalibRecHitRecChi2Algo.h.

31 { };

template<class C >

EcalUncalibRecHitRecChi2Algo< C >::EcalUncalibRecHitRecChi2Algo	(	const C &	dataFrame,
		const double	amplitude,
		const EcalTimeCalibConstant &	timeIC,
		const double	amplitudeOutOfTime,
		const double	jitter,
		const double *	pedestals,
		const double *	pedestalsRMS,
		const double *	gainRatios,
		const EcalShapeBase &	testbeamPulseShape,
		const std::vector< double > &	chi2Parameters
	)

Definition at line 57 of file EcalUncalibRecHitRecChi2Algo.h.

References ecalMGPA::gainId(), CastorSimpleRecAlgoImpl::isSaturated(), cmsBatch::log, EcalShapeBase::timeToRise(), and tzero.

 {
 
     double noise_A = chi2Parameters[0]; // noise term for in-time chi2
     double const_A = chi2Parameters[1]; // constant term for in-time chi2
     double noise_B = chi2Parameters[2]; // noise term for out-of-time chi2
     double const_B = chi2Parameters[3]; // constant term for out-of-time chi2
 
 
     chi2_=0;
     chi2OutOfTime_=0;
     double S_0=0;  // will store the first mgpa sample
     double ped_ave=0;  // will store the average pedestal
 
     int gainId0 = 1;
     int iGainSwitch = 0;
     bool isSaturated = 0;
     for(int iSample = 0; iSample < C::MAXSAMPLES; iSample++) // if gain switch use later the pedestal RMS, otherwise we use the pedestal from the DB
     {
       int gainId = dataFrame.sample(iSample).gainId();
       if(gainId == 0)
       {
     gainId = 3; // if saturated, treat it as G1
     isSaturated = 1;
       }
       if(gainId != gainId0)iGainSwitch = 1;
 
       if(gainId==1 && iSample==0)S_0 = dataFrame.sample(iSample).adc(); // take only first presample to estimate the pedestal
       if(gainId==1 && iSample<3)ped_ave += (1/3.0)*dataFrame.sample(iSample).adc(); // take first 3 presamples to estimate the pedestal
     }
 
 
     // compute testbeamPulseShape shape parameters
     double ADC_clock = 25; // 25 ns
     double risingTime = testbeamPulseShape.timeToRise();
     double tzero = risingTime  - 5*ADC_clock;  // 5 samples before the peak
 
     double shiftTime = + timeIC; // we put positive here
     double shiftTimeOutOfTime = -jitter*ADC_clock; // we put negative here
     
 
     bool readoutError = false;
 
     for(int iSample = 0; iSample < C::MAXSAMPLES; iSample++)
     {
         int gainId = dataFrame.sample(iSample).gainId();
     if(dataFrame.sample(iSample).adc()==0)readoutError=true;
         if(gainId==0)continue; // skip saturated samples
 
 
         double ped = !iGainSwitch ? ped_ave:pedestals[gainId-1];  // use dynamic pedestal for G12 and average pedestal for G6,G1
     //double pedRMS = pedestalsRMS[gainId-1];
         double S_i = double(dataFrame.sample(iSample).adc());
 
     // --- calculate in-time chi2
 
 
         double f_i = (testbeamPulseShape)(tzero + shiftTime + iSample*ADC_clock);
     double R_i = (S_i- ped)*gainRatios[gainId-1] - f_i*amplitude;
     double R_iErrorSquare = noise_A*noise_A + const_A*const_A*amplitude*amplitude;
 
     chi2_ += R_i*R_i/R_iErrorSquare;
 
     // --- calculate out-of-time chi2
 
         double g_i = (testbeamPulseShape)(tzero + shiftTimeOutOfTime + iSample*ADC_clock); // calculate out of time chi2
 
         double R_iOutOfTime = (S_i- S_0)*gainRatios[gainId-1] - g_i*amplitudeOutOfTime;
     double R_iOutOfTimeErrorSquare = noise_B*noise_B + const_B*const_B*amplitudeOutOfTime*amplitudeOutOfTime;
 
  
         chi2OutOfTime_ += R_iOutOfTime*R_iOutOfTime/R_iOutOfTimeErrorSquare;
     }
 
 
     if(!isSaturated && !iGainSwitch && chi2_>0 && chi2OutOfTime_>0) 
     {   
     chi2_ = 7*(3+log(chi2_)); chi2_ = chi2_<0 ? 0:chi2_; // this is just a convinient mapping for storing in the calibRecHit bit map 
     chi2OutOfTime_ = 7*(3+log(chi2OutOfTime_)); chi2OutOfTime_ = chi2OutOfTime_<0 ? 0:chi2OutOfTime_; 
     }else
     {
         chi2_=0;
     chi2OutOfTime_=0;
     }
 
     if(readoutError) // rare situation
     {
     chi2_=99.0;  // chi2 is very large in these cases, put a code value to discriminate against normal noise
     chi2OutOfTime_=99.0;
     }
 }