CMS 3D CMS Logo

EcalUncalibRecHitFixedAlphaBetaAlgo.h
Go to the documentation of this file.
1 #ifndef RecoLocalCalo_EcalRecAlgos_EcalUncalibRecHitFixedAlphaBetaAlgo_HH
2 #define RecoLocalCalo_EcalRecAlgos_EcalUncalibRecHitFixedAlphaBetaAlgo_HH
3 
14 #include "CLHEP/Matrix/Vector.h"
15 #include "CLHEP/Matrix/SymMatrix.h"
16 //#include "CLHEP/Matrix/Matrix.h"
18 
21 
22 #include <vector>
23 #include <string>
24 
25 //#include "TROOT.h"
26 //#include "TMinuit.h"
27 //#include "TGraph.h"
28 //#include "TF1.h"
29 //#include "TMatrixD.h"
30 //#include "TVectorD.h"
31 
33 {
34 
35 
36  private:
37  double MinAmpl_;
39  double fAlpha_;//parameter of the shape
40  double fBeta_;//parameter of the shape
41  double fAmp_max_;// peak amplitude
42  double fTim_max_ ;// time of the peak (in 25ns units)
43  double fPed_max_;// pedestal value
44  double alfabeta_;
45 
46 
47  int fNb_iter_;
50 
51  float fSigma_ped;
52  double un_sur_sigma;
53  //temporary solution for different alpha and beta
54  float alpha_table_[36][1701];
55  float beta_table_[36][1701];
56 
57  bool doFit_;
58 
59  double pulseShapeFunction(double t);
60  float PerformAnalyticFit(double* samples, int max_sample);
61  void InitFitParameters(double* samples, int max_sample);
62  CLHEP::HepSymMatrix DM1_ ; CLHEP::HepVector temp_;
63  public:
64 
66  un_sur_sigma = 1./double(fSigma_ped) ;
67  for (int i=0;i<36;i++){
68  for(int j=0;j<1701;j++){
69  alpha_table_[i][j] = 1.2 ;
70  beta_table_[i][j] = 1.7 ;
71  }
72  }
73  doFit_ = false;
74  MinAmpl_ = 16;
75  dyn_pedestal = true;
76  }
77  EcalUncalibRecHitFixedAlphaBetaAlgo<C>(int n_iter, int n_bef_max =1, int n_aft_max =3, float sigma_ped = 1.1):fAlpha_(0.),fBeta_(0.),fAmp_max_(-1.),fTim_max_(-1),fPed_max_(0),alfabeta_(0),DM1_(3),temp_(3){
78 
79  fNb_iter_ = n_iter ;
80  fNum_samp_bef_max_ = n_bef_max ;
81  fNum_samp_after_max_ = n_aft_max ;
82  fSigma_ped = sigma_ped;
83  un_sur_sigma = 1./double(fSigma_ped) ;
84  for (int i=0;i<36;i++){
85  for(int j=0;j<1701;j++){
86  alpha_table_[i][j] = 1.2 ;
87  beta_table_[i][j] = 1.7 ;
88  }
89  }
90  doFit_ = false;
91  MinAmpl_ = 16;
92  dyn_pedestal = true;
93  };
94 
96  EcalUncalibratedRecHit makeRecHit(const C& dataFrame, const double* pedestals,
97  const double* gainRatios,
99  const EcalWeightSet::EcalChi2WeightMatrix** chi2Matrix) override;
100  void SetAlphaBeta( double alpha, double beta);
101  void SetMinAmpl(double ampl);
102  void SetDynamicPedestal(bool dyn_pede);
103 };
104 
105 
106 
108 template<class C> EcalUncalibratedRecHit EcalUncalibRecHitFixedAlphaBetaAlgo<C>::makeRecHit(const C& dataFrame, const double* pedestals,
109  const double* gainRatios,
111  const EcalWeightSet::EcalChi2WeightMatrix** chi2Matrix){
112  double chi2_(-1.);
113 
114  // double Gain12Equivalent[4]={0,1,2,12};
115  double frame[C::MAXSAMPLES];// will contain the ADC values
116  double pedestal =0; // carries pedestal for gain12 i.e. gainId==1
117 
118  int gainId0 = 1; // expected gainId at the beginning of dataFrame
119  int iGainSwitch = 0; // flags whether there's any gainId other than gainId0
120  int GainId= 0; // stores gainId at every sample
121  double maxsample(-1); // ADC value of maximal ped-subtracted sample
122  int imax(-1); // sample number of maximal ped-subtracted sample
123  bool external_pede = false;
124  bool isSaturated = false; // flag reporting whether gain0 has been found
125 
126  // Get time samples checking for Gain Switch and pedestals
127  if(pedestals){
128  external_pede = true;
129  if(dyn_pedestal) { pedestal = (double(dataFrame.sample(0).adc()) + double(dataFrame.sample(1).adc()))/2.;}
130  else{ pedestal = pedestals[0];}
131  for(int iSample = 0; iSample < C::MAXSAMPLES; iSample++) {
132  //create frame in adc gain 12 equivalent
133  GainId = dataFrame.sample(iSample).gainId();
134 
135  // FIX-ME: warning: the vector pedestal is supposed to have in the order G12, G6 and G1
136  // if GainId is zero treat it as 3 temporarily to protect against undefined
137  // frame will be set to ~max of gain1
138  if ( GainId == 0 )
139  {
140  GainId = 3;
141  isSaturated = true;
142  }
143 
144  if (GainId != gainId0) iGainSwitch = 1;
145 
146  if(GainId==gainId0){frame[iSample] = double(dataFrame.sample(iSample).adc())-pedestal ;}
147  else {frame[iSample] = (double(dataFrame.sample(iSample).adc())-pedestals[GainId-1])*gainRatios[GainId-1];}
148 
149  if( frame[iSample]>maxsample ) {
150  maxsample = frame[iSample];
151  imax = iSample;
152  }
153  }
154  }
155  else {// pedestal from pre-sample
156  external_pede = false;
157  pedestal = (double(dataFrame.sample(0).adc()) + double(dataFrame.sample(1).adc()))/2.;
158 
159  for(int iSample = 0; iSample < C::MAXSAMPLES; iSample++) {
160  //create frame in adc gain 12 equivalent
161  GainId = dataFrame.sample(iSample).gainId();
162  //no gain switch forseen if there is no external pedestal
163  if ( GainId == 0 )
164  {
165  GainId = 3;
166  isSaturated = true;
167  }
168 
169  frame[iSample] = double(dataFrame.sample(iSample).adc())-pedestal ;
170  // if gain has switched but no pedestals are available, no much good you can do...
171  if (GainId > gainId0) iGainSwitch = 1;
172  if( frame[iSample]>maxsample ) {
173  maxsample = frame[iSample];
174  imax = iSample;
175  }
176  }
177  }
178 
179  if( (iGainSwitch==1 && external_pede==false) || // ... thus you return dummy rechit
180  imax ==-1 ){ // protect against all frames being <-1
181  return EcalUncalibratedRecHit( dataFrame.id(), -1., -100., -1. , -1.);
182  }
183 
184  InitFitParameters(frame, imax);
185  chi2_ = PerformAnalyticFit(frame,imax);
186  uint32_t flags = 0;
187  if (isSaturated) flags = EcalUncalibratedRecHit::kSaturated;
188 
189  /* std::cout << "separate fits\nA: " << fAmp_max_ << ", ResidualPed: " << fPed_max_
190  <<", pedestal: "<<pedestal << ", tPeak " << fTim_max_ << std::endl;
191  */
192  return EcalUncalibratedRecHit( dataFrame.id(),fAmp_max_, pedestal+fPed_max_, fTim_max_ - 5 , chi2_, flags );
193 }
194 
196  if( alfabeta_ <= 0 ) return((double)0.);
197  double dtsbeta,variable,puiss;
198  double dt = t-fTim_max_ ;
199  if(dt > -alfabeta_) {
200  dtsbeta=dt/fBeta_ ;
201  variable=1.+dt/alfabeta_ ;
202  puiss=pow(variable,fAlpha_);
203  return fAmp_max_*puiss*exp(-dtsbeta)+fPed_max_ ;
204  }
205  return fPed_max_ ;
206 }
207 
208 template<class C> void EcalUncalibRecHitFixedAlphaBetaAlgo<C>::InitFitParameters(double* samples, int max_sample){
209 
210  // in a first attempt just use the value of the maximum sample
211  fAmp_max_ = samples[max_sample];
212  fTim_max_ = max_sample;
213  fPed_max_ = 0;
214 
215  // amplitude too low for fit to converge
216  // timing set correctly is assumed
217  if(fAmp_max_ < MinAmpl_){
218  fAmp_max_ = samples[5];
219  double sumA = samples[5]+samples[4]+samples[6];
220  if(sumA != 0) { fTim_max_ = 5+(samples[6]-samples[4])/sumA; }
221  else{ fTim_max_ = -993; }//-999+6
222  doFit_ = false;
223  }
224  // if timing very badly off, that just use max sample
225  else if(max_sample <1 || max_sample > 7)
226  { doFit_ = false;}
227  else{
228  //y=a*(x-xM)^2+b*(x-xM)+c
229  doFit_ = true;
230  float a = float(samples[max_sample-1]+samples[max_sample+1]-2*samples[max_sample])/2.;
231  if(a==0){doFit_ =false; return;}
232  float b = float(samples[max_sample+1]-samples[max_sample-1])/2.;
233 
234  fTim_max_ = max_sample - b/(2*a);
235  fAmp_max_ = samples[max_sample] - b*b/(4*a);
236  }
237 
238 }
239 
240 template<class C> float EcalUncalibRecHitFixedAlphaBetaAlgo<C>::PerformAnalyticFit(double* samples, int max_sample){
241 
242  //int fValue_tim_max = max_sample;
247  //| the pedestal (fPed_max_)
248 
249  double chi2=-1 , db[3] ;
250 
251 
252  //HepSymMatrix DM1(3) ; CLHEP::HepVector temp(3) ;
253 
254  int num_fit_min =(int)(max_sample - fNum_samp_bef_max_ ) ;
255  int num_fit_max =(int)(max_sample + fNum_samp_after_max_) ;
256 
257  if (num_fit_min<0) num_fit_min = 0 ;
258  //if (num_fit_max>=fNsamples-1) num_fit_max = fNsamples-2 ;
259  if (num_fit_max>= C::MAXSAMPLES ) {num_fit_max = C::MAXSAMPLES-1 ;}
260 
261  if(! doFit_ ) {
262  LogDebug("EcalUncalibRecHitFixedAlphaBetaAlgo")<<"No fit performed. The amplitude of sample 5 will be used";
263  return -1;
264  }
265 
266  double func,delta ;
267  double variation_func_max = 0. ;double variation_tim_max = 0. ; double variation_ped_max = 0. ;
269  for (int iter=0 ; iter < fNb_iter_ ; iter ++) {
271  chi2 = 0. ; //PROD.Zero() ; DM1.Zero() ;
272 
273  for(int i1=0 ; i1<3 ; i1++) {
274  temp_[i1]=0;
275  for(int j1=i1 ; j1<3 ; j1++) {
276  DM1_.fast(j1+1,i1+1) = 0; }
277  }
278 
279  fAmp_max_ += variation_func_max ;
280  fTim_max_ += variation_tim_max ;
281  fPed_max_ += variation_ped_max ;
282 
284  for( int i = num_fit_min ; i <= num_fit_max ; i++) {
285  //if(i>fsamp_edge_fit && i<num_fit_min) continue ; // remove front edge samples
287  func = pulseShapeFunction( (double)i ) ;
289  double dt =(double)i - fTim_max_ ;
290  if(dt > -alfabeta_) {
291  double dt_sur_beta = dt/fBeta_ ;
292  double variable = (double)1. + dt/alfabeta_ ;
293  double expo = exp(-dt_sur_beta) ;
294  double puissance = pow(variable,fAlpha_) ;
295 
296  db[0]=un_sur_sigma*puissance*expo ;
297  db[1]=fAmp_max_*db[0]*dt_sur_beta/(alfabeta_*variable) ;
298  }
299  else {
300  db[0]=0. ; db[1]=0. ;
301  }
302  db[2]=un_sur_sigma ;
304  for(int i1=0 ; i1<3 ; i1++) {
305  for(int j1=i1 ; j1<3 ; j1++) {
306  //double & fast(int row, int col);
307  DM1_.fast(j1+1,i1+1) += db[i1]*db[j1]; }
308  }
310  delta = (samples[i]-func)*un_sur_sigma ;
312  for(int ii=0 ; ii<3 ;ii++) {temp_[ii] += delta*db[ii] ;}
313  chi2 += delta * delta ;
314  }
315 
316  int fail=0 ;
317  DM1_.invert(fail) ;
318  if(fail != 0.) {
319  //just a guess from the value of the parameters in the previous interaction;
320  //printf("wH4PulseFitWithFunction =====> determinant error --> No Fit Provided !\n") ;
321  InitFitParameters(samples,max_sample);
322  return -101 ;
323  }
324 /* for(int i1=0 ; i1<3 ; i1++) { */
325 /* for(int j1=0 ; j1<3 ; j1++) { */
326 /* //double & fast(int row, int col); */
327 /* std::cout<<"inverted: "<<DM1[j1][i1]<<std::endl;;} */
328 /* } */
329 /* std::cout<<"vector temp: "<< temp[0]<<" "<<temp[1]<<" "<<temp[2]<<std::endl; */
331  CLHEP::HepVector PROD = DM1_*temp_ ;
332  // std::cout<<"vector PROD: "<< PROD[0]<<" "<<PROD[1]<<" "<<PROD[2]<<std::endl;
333 
334  // Probably the fastest way to protect against
335  // +-inf value in the matrix DM1_ after inversion
336  // (which is nevertheless flagged as successfull...)
337  if ( edm::isNotFinite( PROD[0] ) ) {
338  InitFitParameters(samples,max_sample);
339  return -103 ;
340  }
341 
342  variation_func_max = PROD[0] ;
343  variation_tim_max = PROD[1] ;
344  variation_ped_max = PROD[2] ;
345  //chi2 = chi2/((double)nsamp_used - 3.) ;
346  }
347 
348 
350  if( variation_func_max > 2000. || variation_func_max < -1000. ) {
351  InitFitParameters(samples,max_sample);
352  return -102;
353  }
354 
355 
357  fAmp_max_ += variation_func_max ;
358  fTim_max_ += variation_tim_max ;
359  fPed_max_ += variation_ped_max ;
360 
361 
362  // protection against unphysical results:
363  // ampli mismatched to MaxSample, ampli largely negative, time off preselected range
364  if( fAmp_max_>2*samples[max_sample] || fAmp_max_<-100 || fTim_max_<0 || 9<fTim_max_ ) {
365  InitFitParameters(samples,max_sample);
366  return -104;
367  }
368 
369 
370  //std::cout <<"chi2: "<<chi2<<" ampl: "<<fAmp_max_<<" time: "<<fTim_max_<<" pede: "<<fPed_max_<<std::endl;
371  return chi2;
372 }
373 
374 template<class C> void EcalUncalibRecHitFixedAlphaBetaAlgo<C>::SetAlphaBeta( double alpha, double beta){
375  fAlpha_ = alpha;
376  fBeta_= beta;
378 }
379 
380 template<class C> void EcalUncalibRecHitFixedAlphaBetaAlgo<C>::SetMinAmpl( double ampl){
381  MinAmpl_ = ampl;
382 }
384  dyn_pedestal = p;
385 }
386 #endif
#define LogDebug(id)
const double beta
dbl * delta
Definition: mlp_gen.cc:36
void InitFitParameters(double *samples, int max_sample)
float dt
Definition: AMPTWrapper.h:126
float alpha
Definition: AMPTWrapper.h:95
std::vector< Variable::Flags > flags
Definition: MVATrainer.cc:135
EcalUncalibratedRecHit makeRecHit(const C &dataFrame, const double *pedestals, const double *gainRatios, const EcalWeightSet::EcalWeightMatrix **weights, const EcalWeightSet::EcalChi2WeightMatrix **chi2Matrix) override
Compute parameters.
bool isNotFinite(T x)
Definition: isFinite.h:10
ii
Definition: cuy.py:588
math::Matrix< 10, 10 >::type EcalChi2WeightMatrix
Definition: EcalWeightSet.h:23
double b
Definition: hdecay.h:120
bool isSaturated(const Digi &digi, const int &maxADCvalue, int ifirst, int n)
double a
Definition: hdecay.h:121
#define PROD(A, B)
math::Matrix< 3, 10 >::type EcalWeightMatrix
Definition: EcalWeightSet.h:22
float PerformAnalyticFit(double *samples, int max_sample)
def fail(errstr="")
Power< A, B >::type pow(const A &a, const B &b)
Definition: Power.h:40