#include <PulseFitWithFunction.h>

Inheritance diagram for PulseFitWithFunction:

Public Member Functions
virtual double	doFit (double *)

double	getAmpl ()

double	getAmpl_parab ()

double	getMax_parab ()

int	getSampMax_parab ()

double	getTime ()

double	getTime_parab ()

virtual void	init (int, int, int, int, double, double)

	PulseFitWithFunction ()

virtual	~PulseFitWithFunction ()

Public Attributes
double	fAmp_fitted_max

double	fFunc_max

int	fNumber_samp_max

double	fSigma_ped

double	fTim_fitted_max

double	fTim_max

double	fValue_tim_max

Private Member Functions
double	Electronic_shape (double)

double	Fit_electronic (int, double *, double)

void	Fit_parab (double , int, int, double )

Private Attributes
double	amp_max

double	amp_parab

double	fAlpha

double	fAlpha_beam

double	fAlpha_laser

double	fBeta

double	fBeta_beam

double	fBeta_laser

int	fNb_iter

int	fNsamples

int	fNum_samp_after_max

int	fNum_samp_bef_max

int	imax

double	tim_parab

Detailed Description

Definition at line 26 of file PulseFitWithFunction.h.

Constructor & Destructor Documentation

PulseFitWithFunction::PulseFitWithFunction ( )

Definition at line 28 of file PulseFitWithFunction.cc.

References fNsamples, fNum_samp_after_max, and fNum_samp_bef_max.

 { 
  
   fNsamples               =  0;
   fNum_samp_bef_max       =  0;
   fNum_samp_after_max     =  0;
 }

PulseFitWithFunction::~PulseFitWithFunction ( )

virtual

Definition at line 37 of file PulseFitWithFunction.cc.

38 {

39 }

Member Function Documentation

double PulseFitWithFunction::doFit ( double * adc )

virtual

Definition at line 68 of file PulseFitWithFunction.cc.

References amp_max, amp_parab, HLT_FULL_cff::chi2, fFunc_max, Fit_electronic(), Fit_parab(), fNsamples, fNumber_samp_max, fTim_max, fValue_tim_max, imax, and tim_parab.

 {
   double parout[4]; // amp_max ;
     //double amp_parab , tim_parab ;
   double chi2;
   //int imax ;
 //
 // first  one has to get starting point first with parabolic fun// //  
   Fit_parab(&adc[0],3,fNsamples,parout) ;
   amp_parab = parout[0] ;
   tim_parab = parout[1] ;
   imax = (int)parout[2] ;
   amp_max = parout[3] ;
   //printf("amp_parab= %f tim_parab=%f amp_max=%f imax=%d\n",amp_parab,tim_parab,amp_max,imax); 
   fNumber_samp_max=imax ;
 //
   
   if(amp_parab < 1.) {
      tim_parab = (double)imax ;
       amp_parab = amp_max ;
   }
 //
   fValue_tim_max= tim_parab ;
   fFunc_max= amp_parab ;
   fTim_max=tim_parab ;
 
 //  here to fit maximum amplitude and time of arrival ...
 
   chi2 = Fit_electronic(0,&adc[0],8.) ;
                                    // adc is an array to be filled with samples
                                    // 0 is for Laser (1 for electron)  
                                    // 8 is for sigma of pedestals 
                                    // which (can be computed)
 
 
   //  double amplitude = fAmp_fitted_max ; // amplitude fitted 
   // double time = fTim_fitted_max ; // time fitted
 
  
 
   return chi2 ; // return amplitude fitted
 
 }

double PulseFitWithFunction::Electronic_shape ( double tim )

private

Definition at line 226 of file PulseFitWithFunction.cc.

References dt, fAlpha, fBeta, fFunc_max, and fTim_max.

Referenced by Fit_electronic().

 {
   // electronic function (from simulation) to fit ECAL pulse shape
   double func_electronic,dtsbeta,variable,puiss;
   double albet = fAlpha*fBeta ;
   if( albet <= 0 ) return( (Double_t)0. );
   double dt = tim-fTim_max ;
   if(dt > -albet)  {
     dtsbeta=dt/fBeta ;
     variable=1.+dt/albet ;
     puiss=TMath::Power(variable,fAlpha);
     func_electronic=fFunc_max*puiss*TMath::Exp(-dtsbeta);
   }
   else func_electronic = 0. ;
   //
   return func_electronic ;
 }

double PulseFitWithFunction::Fit_electronic	(	int	data,
		double *	adc_to_fit,
		double	sigmas_sample
	)

private

Definition at line 116 of file PulseFitWithFunction.cc.

References HLT_FULL_cff::chi2, delta, dt, Electronic_shape(), fAlpha, fAlpha_beam, fAlpha_laser, fAmp_fitted_max, fBeta, fBeta_beam, fBeta_laser, fFunc_max, fNb_iter, fNum_samp_after_max, fNum_samp_bef_max, fNumber_samp_max, fSigma_ped, fTim_fitted_max, fTim_max, fValue_tim_max, and i.

Referenced by doFit().

                                                                                                 {
   // fit electronic function from simulation
   // parameters fAlpha and fBeta are fixed and fit is providing
   // the maximum amplitude ( fAmp_fitted_max ) and the time of
   // the maximum amplitude ( fTim_fitted_max) 
   // initialization of parameters
   double chi2=0;
   double d_alpha, d_beta ;
   // first initialize parameters fAlpha and fBeta ( depending of beam or laser)
   fAlpha = fAlpha_laser ;
   fBeta = fBeta_laser ;
   if(data == 1) { 
     fAlpha = fAlpha_beam ;
     fBeta = fBeta_beam ;
   }
   //
   fAmp_fitted_max = 0. ;
   fTim_fitted_max = 0. ;
   double un_sur_sigma = 1./fSigma_ped ;
   double variation_func_max = 0. ;
   double variation_tim_max = 0. ;
   //
   if(fValue_tim_max > 20. || fValue_tim_max  < 3.) {
       fValue_tim_max = fNumber_samp_max ;
   }
   int num_fit_min =(int)(fValue_tim_max - fNum_samp_bef_max) ;
   int num_fit_max =(int)(fValue_tim_max + fNum_samp_after_max) ;
   //
 
   if( sigmas_sample > 0. ) un_sur_sigma = 1./sigmas_sample;
   else un_sur_sigma = 1.;
 
   double func,delta ;
   //          Loop on iterations        
   for (int iter=0 ; iter < fNb_iter ; iter ++) {
    
     //          initialization inside iteration loop !
     chi2 = 0. ;
     double d11 = 0. ;
     double d12 = 0. ;
     double d22 = 0. ;
     double z1 = 0. ;
     double z2 = 0. ;
     fFunc_max += variation_func_max ;
     fTim_max += variation_tim_max ;
     int nsamp_used = 0 ;
     //
    
     // Then we loop on samples to be fitted
 
 
     for( int i = num_fit_min ; i < num_fit_max+1 ; i++) {
       // calculate function to be fitted
      
       func = Electronic_shape( (double)i  ) ;
       
       // then calculate derivatives of function to be fitted
       double dt =(double)i - fTim_max ;
       double alpha_beta = fAlpha*fBeta  ;
      
      if(dt > -alpha_beta)  {      
       double dt_sur_beta = dt/fBeta ;
      
       double variable = (double)1. + dt/alpha_beta ;
       double expo = TMath::Exp(-dt_sur_beta) ;   
      
       double puissance = TMath::Power(variable,fAlpha) ;
       d_alpha=un_sur_sigma*puissance*expo ;
       d_beta=fFunc_max*d_alpha*dt_sur_beta/(alpha_beta*variable) ;
      }
       else {
     continue ;
      }
      
       nsamp_used ++ ; // number of samples used inside the fit
       // compute matrix elements D (symetric --> d12 = d21 )
       d11 += d_alpha*d_alpha ;
       d12 += d_alpha*d_beta ;
       d22 += d_beta*d_beta ;
       // compute delta
       delta = (adc_to_fit[i]-func)*un_sur_sigma ;
       // compute vector elements Z
       z1 += delta*d_alpha ;
       z2 += delta*d_beta ;
       chi2 += delta * delta ;
     } //             end of loop on samples  
     double denom = d11*d22-d12*d12 ;
     if(denom == 0.) {
       //printf( "attention denom = 0 signal pas fitte \n") ;
       return 101 ;
     }
     if(nsamp_used < 3) {
       //printf( "Attention nsamp = %d ---> no function fit provided \n",nsamp_used) ;
       return 102 ;
     }
     // compute variations of parameters fAmp_max and fTim_max 
     variation_func_max = (z1*d22-z2*d12)/denom ;
     variation_tim_max = (-z1*d12+z2*d11)/denom ;
     chi2 = chi2/((double)nsamp_used - 2.) ;
   } //      end of loop on iterations       
   // results of the fit are calculated 
   fAmp_fitted_max = fFunc_max + variation_func_max ;
   fTim_fitted_max = fTim_max + variation_tim_max ;
   //
   return chi2 ;
 }

void PulseFitWithFunction::Fit_parab	(	double *	,
		int	,
		int	,
		double *
	)

private

Definition at line 243 of file PulseFitWithFunction.cc.

References dt, imax, relval_2017::k, and reco::return().

Referenced by doFit().

 {
 /* Now we calculate the parabolic adjustement in order to get        */
 /*    maximum and time max                                           */
  
   double denom,dt,amp1,amp2,amp3 ;
   double ampmax = 0. ;              
   int imax = 0 ;
   int k ;
 /*
                                                                    */     
   for ( k = nmin ; k < nmax ; k++) {
     //printf("ampl[%d]=%f\n",k,ampl[k]);
     if (ampl[k] > ampmax ) {
       ampmax = ampl[k] ;
       imax = k ;
     }
   }
     amp1 = ampl[imax-1] ;
     amp2 = ampl[imax] ;
     amp3 = ampl[imax+1] ;
     denom=2.*amp2-amp1-amp3  ;
 /*                                       */       
     if(denom>0.){
       dt =0.5*(amp3-amp1)/denom  ;
     }
     else {
       //printf("denom =%f\n",denom)  ;
       dt=0.5  ;
     }
 /*                                           */        
 /* ampmax correspond au maximum d'amplitude parabolique et dt        */
 /* decalage en temps par rapport au sample maximum soit k + dt       */
         
     parout[0] =amp2+(amp3-amp1)*dt*0.25 ;
     parout[1] = (double)imax + dt ;
         parout[2] = (double)imax ;
     parout[3] = ampmax ;
 return ;
 }

double PulseFitWithFunction::getAmpl ( )

inline

Definition at line 53 of file PulseFitWithFunction.h.

References fAmp_fitted_max.

53 { return fAmp_fitted_max; }

PulseFitWithFunction::fAmp_fitted_max

double fAmp_fitted_max

Definition: PulseFitWithFunction.h:44

double PulseFitWithFunction::getAmpl_parab ( )

inline

Definition at line 50 of file PulseFitWithFunction.h.

References amp_parab.

50 { return amp_parab; }

PulseFitWithFunction::amp_parab

double amp_parab

Definition: PulseFitWithFunction.h:61

double PulseFitWithFunction::getMax_parab ( )

inline

Definition at line 56 of file PulseFitWithFunction.h.

References amp_max.

56 { return amp_max; }

PulseFitWithFunction::amp_max

double amp_max

Definition: PulseFitWithFunction.h:61

int PulseFitWithFunction::getSampMax_parab ( )

inline

Definition at line 57 of file PulseFitWithFunction.h.

References imax.

57 { return imax; }

PulseFitWithFunction::imax

int imax

Definition: PulseFitWithFunction.h:62

double PulseFitWithFunction::getTime ( )

inline

Definition at line 54 of file PulseFitWithFunction.h.

References fTim_fitted_max.

54 { return fTim_fitted_max; }

PulseFitWithFunction::fTim_fitted_max

double fTim_fitted_max

Definition: PulseFitWithFunction.h:45

double PulseFitWithFunction::getTime_parab ( )

inline

Definition at line 51 of file PulseFitWithFunction.h.

References tim_parab.

51 { return tim_parab; }

PulseFitWithFunction::tim_parab

double tim_parab

Definition: PulseFitWithFunction.h:61

void PulseFitWithFunction::init	(	int	n_samples,
		int	samplb,
		int	sampla,
		int	niter,
		double	alfa,
		double	beta
	)

virtual

Definition at line 43 of file PulseFitWithFunction.cc.

References beta, fAlpha_beam, fAlpha_laser, fBeta_beam, fBeta_laser, fNb_iter, fNsamples, fNum_samp_after_max, fNum_samp_bef_max, and reco::return().

 {
   //printf("\n");
   //printf(" =========================================================\n");
   //printf(" ==     Initialising the function fit method            ==\n");
   //printf(" ==          PulseFitWithFunction::init                 ==\n");
   //printf(" ==                                                     ==\n");
 
   
   fNsamples   = n_samples ;
   fAlpha_laser = alfa;
   fBeta_laser = beta ;
   fAlpha_beam = 0.98 ;
   fBeta_beam = 2.04 ;
   fNb_iter = niter ;
   fNum_samp_bef_max = samplb ;
   fNum_samp_after_max = sampla  ;
   //printf(" ==  # samples used = %3d                               ==\n",fNsamples);
   //printf(" ==  #sample before max= %1d and #sample after maximum= %1d ==\n",fNum_samp_bef_max,fNum_samp_after_max);
   //printf(" ==           alpha= %5.4f       beta= %5.4f          ==\n",fAlpha_laser,fBeta_laser);
   //printf(" =========================================================\n\n");
   return ;
  }