+// -*- mode: c++ -*-
/**************************************************************************
* This file is property of and copyright by *
* the Relativistic Heavy Ion Group (RHIG), Yale University, US, 2009 *
#include <iostream>
using namespace std;
-//ClassImp(AliCaloRawAnalyzer)
+ClassImp(AliCaloRawAnalyzer)
AliCaloRawAnalyzer::AliCaloRawAnalyzer(const char *name, const char *nameshort) : TObject(),
- fMinTimeIndex(-1),
- fMaxTimeIndex(-1),
- fFitArrayCut(5),
- fAmpCut(4),
- fNsampleCut(5),
- fNsamplePed(3),
- fIsZerosupressed( false ),
- fVerbose( false )
+ fMinTimeIndex(-1),
+ fMaxTimeIndex(-1),
+ fFitArrayCut(5),
+ fAmpCut(4),
+ fNsampleCut(5),
+ fOverflowCut(950),
+ fNsamplePed(3),
+ fIsZerosupressed( false ),
+ fVerbose( false ),
+ fAlgo(Algo::kNONE)
{
//Comment
- sprintf(fName, "%s", name);
- sprintf(fNameShort, "%s", nameshort);
+ snprintf(fName, 256,"%s", name);
+ snprintf(fNameShort,256, "%s", nameshort);
for(int i=0; i < MAXSAMPLES; i++ )
{
{
//Selection of subset of data from one bunch that will be used for fitting or
//Peak finding. Go to the left and right of index of the maximum time bin
- //Untile the ADC value is less that fFitArrayCut
+ //Until the ADC value is less that fFitArrayCut, or derivative changes sign (data jump)
int tmpfirst = maxindex;
int tmplast = maxindex;
-
- while(( tmpfirst ) > 0 && ( fData[tmpfirst] > fFitArrayCut ))
+ Double_t prevFirst = fData[maxindex];
+ Double_t prevLast = fData[maxindex];
+ bool firstJump = false;
+ bool lastJump = false;
+
+ while( (tmpfirst >= 0) && (fData[tmpfirst] >= fFitArrayCut) && (!firstJump) )
{
+ // jump check:
+ if (tmpfirst != maxindex) { // neighbor to maxindex can share peak with maxindex
+ if (fData[tmpfirst] >= prevFirst) {
+ firstJump = true;
+ }
+ }
+ prevFirst = fData[tmpfirst];
tmpfirst -- ;
}
- while(( tmplast ) < (length-1) && ( fData [tmplast] > fFitArrayCut ))
+ while( (tmplast < length) && (fData[tmplast] >= fFitArrayCut) && (!lastJump) )
{
+ // jump check:
+ if (tmplast != maxindex) { // neighbor to maxindex can share peak with maxindex
+ if (fData[tmplast] >= prevLast) {
+ lastJump = true;
+ }
+ }
+ prevLast = fData[tmplast];
tmplast ++;
}
+ // we keep one pre- or post- sample if we can (as in online)
+ // check though if we ended up on a 'jump', or out of bounds: if so, back up
+ if (firstJump || tmpfirst<0) tmpfirst ++;
+ if (lastJump || tmplast>=length) tmplast --;
- *first = tmpfirst +1;
- *last = tmplast -1;
-
- int nsamples = *last - *first + 1;
- if (nsamples < fNsampleCut) {
- // keep edge bins (below threshold) also, for low # of samples case
- *first = tmpfirst;
- *last = tmplast;
- }
-
+ *first = tmpfirst;
+ *last = tmplast;
+ return;
}
}
-AliCaloFitResults
-AliCaloRawAnalyzer::Evaluate( const vector<AliCaloBunchInfo> &/*bunchvector*/, const UInt_t /*altrocfg1*/, const UInt_t /*altrocfg2*/)
-{ // method to do the selection of what should possibly be fitted
- // not implemented for base class
- return AliCaloFitResults( 0, 0 );
+Double_t
+AliCaloRawAnalyzer::CalculateChi2(const Double_t amp, const Double_t time,
+ const Int_t first, const Int_t last,
+ const Double_t adcErr,
+ const Double_t tau)
+{
+ // Input:
+ // amp - max amplitude;
+ // time - time of max amplitude;
+ // first, last - sample array indices to be used
+ // adcErr - nominal error of amplitude measurement (one value for all channels)
+ // if adcErr<0 that mean adcErr=1.
+ // tau - filter time response (in timebin units)
+ // Output:
+ // chi2 - chi2
+
+ if (first == last || first<0 ) { // signal consists of single sample, chi2 estimate (0) not too well defined..
+ // or, first is negative, the indices are not valid
+ return Ret::kDummy;
+ }
+
+ int nsamples = last - first + 1;
+ // printf(" AliCaloRawAnalyzer::CalculateChi2 : first %i last %i : nsamples %i : amp %3.2f time %3.2f \n", first, last, nsamples, amp, time);
+
+ Int_t x = 0;
+ Double_t chi2 = 0;
+ Double_t dy = 0.0, xx = 0.0, f=0.0;
+
+ for (Int_t i=0; i<nsamples; i++) {
+ x = first + i; // timebin
+ xx = (x - time + tau) / tau; // help variable
+ f = 0;
+ if (xx > 0) {
+ f = amp * xx*xx * TMath::Exp(2 * (1 - xx )) ;
+ }
+ dy = fReversed[x] - f;
+ chi2 += dy*dy;
+ // printf(" AliCaloRawAnalyzer::CalculateChi2 : %i : y %f -> f %f : dy %f \n", i, fReversed[first+i], f, dy);
+ }
+
+ if (adcErr>0.0) { // weight chi2
+ chi2 /= (adcErr*adcErr);
+ }
+ return chi2;
}
+void
+AliCaloRawAnalyzer::CalculateMeanAndRMS(const Int_t first, const Int_t last,
+ Double_t & mean, Double_t & rms)
+{
+ // Input:
+ // first, last - sample array indices to be used
+ // Output:
+ // mean and RMS of samples
+ //
+ // To possibly be used to differentiate good signals from bad before fitting
+ //
+ mean = Ret::kDummy;
+ rms = Ret::kDummy;
+
+ if (first == last || first<0 ) { // signal consists of single sample, chi2 estimate (0) not too well defined..
+ // or, first is negative, the indices are not valid
+ return;
+ }
+
+ int nsamples = last - first + 1;
+ // printf(" AliCaloRawAnalyzer::CalculateMeanAndRMS : first %i last %i : nsamples %i \n", first, last, nsamples);
+
+ int x = 0;
+ Double_t sampleSum = 0; // sum of samples
+ Double_t squaredSampleSum = 0; // sum of samples squared
+
+ for (Int_t i=0; i<nsamples; i++) {
+ x = first + i;
+ sampleSum += fReversed[x];
+ squaredSampleSum += (fReversed[x] * fReversed[x]);
+ }
+
+ mean = sampleSum / nsamples;
+ Double_t squaredMean = squaredSampleSum / nsamples;
+ // The variance (rms squared) is equal to the mean of the squares minus the square of the mean..
+ rms = sqrt(squaredMean - mean*mean);
+
+ return;
+}
+
+
+
+// AliCaloFitResults
+// AliCaloRawAnalyzer::Evaluate( const vector<AliCaloBunchInfo> &/*bunchvector*/, const UInt_t /*altrocfg1*/, const UInt_t /*altrocfg2*/)
+// { // method to do the selection of what should possibly be fitted
+// // not implemented for base class
+// cout << __FILE__ << ":" << __LINE__ << " " << endl;
+
+// return AliCaloFitResults( 0, 0 );
+// }
+
+
+
int
AliCaloRawAnalyzer::PreFitEvaluateSamples( const vector<AliCaloBunchInfo> &bunchvector, const UInt_t altrocfg1, const UInt_t altrocfg2, Int_t & index, Float_t & maxf, short & maxamp, short & maxrev, Float_t & ped, int & first, int & last)
{ // method to do the selection of what should possibly be fitted
index = SelectBunch( bunchvector, &maxampindex, &maxamp ); // select the bunch with the highest amplitude unless any time constraints is set
- if( index >= 0 && maxamp > fAmpCut) // something valid was found, and non-zero amplitude
+ if( index >= 0 && maxamp >= fAmpCut) // something valid was found, and non-zero amplitude
{
// use more convenient numbering and possibly subtract pedestal
ped = ReverseAndSubtractPed( &(bunchvector.at(index)), altrocfg1, altrocfg2, fReversed );
maxf = TMath::MaxElement( bunchvector.at(index).GetLength(), fReversed );
- if ( maxf > fAmpCut ) // possibly significant signal
+ if ( maxf >= fAmpCut ) // possibly significant signal
{
// select array around max to possibly be used in fit
maxrev = maxampindex - bunchvector.at(index).GetStartBin();
git://git.uio.no / u / mrichter / AliRoot.git / blobdiff