Main Page | Modules | Namespace List | Class Hierarchy | Alphabetical List | Class List | Directories | File List | Namespace Members | Class Members | File Members | Related Pages

FitterEM Class Reference

#include <FitterEM.h>

List of all members.

Public Member Functions

 FitterEM ()
 FitterEM (Int_t, CandStripHandleItr)
 ~FitterEM ()
void SetInputParams (Double_t, Double_t *, Double_t, Double_t)
void QuickInput (Double_t, Double_t, Double_t, Double_t, Double_t, Double_t, Int_t)
void SetStepSize (Double_t *)
void SetNSteps (Int_t *)
Double_t PredictEMLoss (Int_t, Int_t, Double_t &)
void DoFit ()
Bool_t Pass ()
Double_t GetChi2 ()
Int_t GetNDF ()
Double_t * GetFittedPars ()
Double_t * GetFittedErr ()
Int_t GetNIterations ()
void DoTests ()
Double_t GetSignTestStat ()
Double_t GetRunTestStat ()
Double_t GetKSTestStat ()

Private Member Functions

Double_t CalculateChi2 ()
void SetUpStripInfo ()
void SetUpFitBox ()

Private Attributes

CandStripHandleItr fDataStpItr
CacherEM * fCache
BinCorrelationEM * fBinCorrel
Int_t fNStp
Double_t fEnergy
Double_t * fVtx
Double_t fdudz
Double_t fdvdz
Int_t fclosestPlaneToVtx
Float_t fclosestZedDiff
Int_t fclosestZedView
Int_t fclosestUStripToVtx
Int_t fclosestVStripToVtx
Float_t fclosestTPosUDiff
Float_t fclosestTPosVDiff
Int_t fNPlanes
Int_t fNStrips
Int_t fMinBoxPlane
Int_t fMaxBoxPlane
Int_t fMinBoxStrip
Int_t fMaxBoxStrip
Double_t fTotMip
Int_t * fPlaneArray
Int_t * fStripArray
Int_t * fViewArray
Double_t * fMipArray
Double_t * fZedArray
Double_t * fTPosArray
Int_t * fBoxPlaneArray
Int_t * fBoxStripArray
Double_t * fMips
Bool_t fPass
Double_t * fFittedPars
Double_t * fFittedErr
Double_t * fStepSize
Int_t * fNSteps
Double_t fChi2Min
Int_t fNDF
Int_t fNiter
Double_t fSignStat
Double_t fBestSignStat
Double_t fRunStat
Double_t fKSStat

Constructor & Destructor Documentation

FitterEM::FitterEM (   ) 
 

Definition at line 29 of file FitterEM.cxx.

References fBinCorrel, fBoxPlaneArray, fBoxStripArray, fCache, fChi2Min, fclosestPlaneToVtx, fclosestTPosUDiff, fclosestTPosVDiff, fclosestUStripToVtx, fclosestVStripToVtx, fclosestZedDiff, fclosestZedView, fdudz, fdvdz, fEnergy, fFittedErr, fFittedPars, fKSStat, fMaxBoxPlane, fMaxBoxStrip, fMinBoxPlane, fMinBoxStrip, fMipArray, fMips, fNDF, fNiter, fNPlanes, fNSteps, fNStp, fNStrips, fPass, fPlaneArray, fRunStat, fSignStat, fStepSize, fStripArray, fTotMip, fTPosArray, fViewArray, fVtx, and fZedArray.

00029                   {
00030   //this constructor is just for testing function
00031   //do not use
00032   fNStp = 1;
00033 
00034   fNPlanes = 0;
00035   fMinBoxPlane = 999;
00036   fMaxBoxPlane = 0;
00037   fNStrips = 0;
00038   fMinBoxStrip = 999;
00039   fMaxBoxStrip = 0;
00040 
00041   fclosestPlaneToVtx = -1;
00042   fclosestZedDiff = 99999;
00043   fclosestZedView = 0;
00044   fclosestUStripToVtx = -1;
00045   fclosestVStripToVtx = -1;
00046   fclosestTPosUDiff = 99999;
00047   fclosestTPosVDiff = 99999;
00048 
00049   fTotMip = 0;
00050   fPlaneArray = new Int_t[fNStp];
00051   fStripArray = new Int_t[fNStp];
00052   fViewArray  = new Int_t[fNStp];
00053   fMipArray   = new Double_t[fNStp];
00054   fZedArray   = new Double_t[fNStp];
00055   fTPosArray  = new Double_t[fNStp];
00056   fBoxPlaneArray = new Int_t[fNStp];
00057   fBoxStripArray = new Int_t[fNStp];
00058   fMips = NULL;
00059 
00060   fEnergy = 0;
00061   fVtx = NULL;
00062   fdudz = 1;
00063   fdvdz = 1;
00064 
00065   fCache = new CacherEM();
00066   fBinCorrel = NULL;
00067 
00068   fPass = true; //start out optimistic
00069   fFittedPars = NULL; //these always constitute the "Current" fit params
00070   fFittedErr = NULL;
00071   fChi2Min = 99999;
00072   fNDF = 0;
00073   fNiter = 0;
00074   fStepSize = NULL;  
00075   fNSteps = NULL;  
00076 
00077   fSignStat = 0;
00078   fRunStat = 0;
00079   fKSStat = 0;
00080 
00081 }

FitterEM::FitterEM (  Int_t  ,
CandStripHandleItr 
) 
 

Definition at line 84 of file FitterEM.cxx.

References fBinCorrel, fBoxPlaneArray, fBoxStripArray, fCache, fChi2Min, fclosestPlaneToVtx, fclosestTPosUDiff, fclosestTPosVDiff, fclosestUStripToVtx, fclosestVStripToVtx, fclosestZedDiff, fclosestZedView, fDataStpItr, fdudz, fdvdz, fEnergy, fFittedErr, fFittedPars, fKSStat, fMaxBoxPlane, fMaxBoxStrip, fMinBoxPlane, fMinBoxStrip, fMipArray, fMips, fNDF, fNiter, fNPlanes, fNSteps, fNStp, fNStrips, fPass, fPlaneArray, fRunStat, fSignStat, fStepSize, fStripArray, fTotMip, fTPosArray, fViewArray, fVtx, fZedArray, and SetUpStripInfo().

00084                                                    {
00085 
00086   fNStp = nstp;
00087   fDataStpItr = itr;
00088 
00089   fNPlanes = 0;
00090   fMinBoxPlane = 999;
00091   fMaxBoxPlane = 0;
00092   fNStrips = 0;
00093   fMinBoxStrip = 999;
00094   fMaxBoxStrip = 0;
00095 
00096   fclosestPlaneToVtx = -1;
00097   fclosestZedDiff = 99999;
00098   fclosestZedView = 0;
00099   fclosestUStripToVtx = -1;
00100   fclosestVStripToVtx = -1;
00101   fclosestTPosUDiff = 99999;
00102   fclosestTPosVDiff = 99999;
00103 
00104   fTotMip = 0;
00105   fPlaneArray = new Int_t[fNStp];
00106   fStripArray = new Int_t[fNStp];
00107   fViewArray  = new Int_t[fNStp];
00108   fMipArray   = new Double_t[fNStp];
00109   fZedArray   = new Double_t[fNStp];
00110   fTPosArray  = new Double_t[fNStp];
00111   fBoxPlaneArray = new Int_t[fNStp];
00112   fBoxStripArray = new Int_t[fNStp];
00113   fMips = NULL;
00114 
00115   SetUpStripInfo();
00116 
00117   fEnergy = 0;
00118   fVtx = NULL;
00119   fdudz = 1;
00120   fdvdz = 1;
00121 
00122   fCache = new CacherEM();
00123   fBinCorrel = NULL;
00124 
00125   fPass = true; //start out optimistic
00126   fFittedPars = NULL; //these always constitute the "Current" fit params
00127   fFittedErr = NULL;
00128   fChi2Min = 99999;
00129   fNDF = 0;
00130   fNiter = 0;
00131   fStepSize = NULL;  
00132   fNSteps = NULL;  
00133 
00134   fSignStat = 0;
00135   fRunStat = 0;
00136   fKSStat = 0;
00137 
00138 }

FitterEM::~FitterEM (   ) 
 

Definition at line 142 of file FitterEM.cxx.

References MSG.

00142                    {
00143   delete [] fVtx;
00144   delete [] fFittedPars;
00145   delete [] fFittedErr;
00146   delete [] fPlaneArray;
00147   delete [] fStripArray;
00148   delete [] fViewArray;
00149   delete [] fMipArray;
00150   delete [] fZedArray;
00151   delete [] fTPosArray;
00152   delete [] fBoxPlaneArray;
00153   delete [] fBoxStripArray;
00154   delete [] fMips;
00155   delete [] fStepSize;
00156   delete [] fNSteps;
00157   delete fCache;
00158   if(fBinCorrel) delete fBinCorrel;
00159   TF2 *prof = NULL; //this should have been created in the course of fitting
00160   if(gROOT->FindObject("FITSHOWEREM_2DPROFILE")) { 
00161     prof = (TF2*) gROOT->FindObject("FITSHOWEREM_2DPROFILE");
00162     delete prof;   //delete it!
00163   }
00164   MSG("FitShowerEM",Msg::kDebug) << "Deleting FitterEM" << endl;
00165 }

Member Function Documentation

Double_t FitterEM::CalculateChi2 (   )  [private]
 

Definition at line 466 of file FitterEM.cxx.

References Chi2(), fBinCorrel, fFittedPars, fMaxBoxPlane, fMaxBoxStrip, fMinBoxPlane, fMinBoxStrip, fMips, fNDF, fNPlanes, fSignStat, BinCorrelationEM::GetErrArray(), MSG, PredictEMLoss(), and BinCorrelationEM::ReInit().

Referenced by DoFit().

00466                                 {
00467   
00468   fBinCorrel->ReInit(fFittedPars);
00469   Double_t *ErrArray = fBinCorrel->GetErrArray(fMinBoxPlane,fMaxBoxPlane,
00470                                                fMinBoxStrip,fMaxBoxStrip);
00471   
00472   Int_t TotStp = fNPlanes*fNStrips;
00473 
00474   //for holding all necessary function values during calculation
00475   Double_t *ff = new Double_t[TotStp];  //fit func values
00476   Double_t *fff = new Double_t[TotStp]; //fit func fluctuations
00477   for(Int_t i=0;i<TotStp;i++) { 
00478     ff[i] = -1; 
00479     fff[i] = 0.; 
00480   }
00481   
00482   Double_t lastDiagTerm = 1;
00483   Double_t flucVal = 0;
00484   //normalise covariance matrix using function: (make ff a percentage)
00485   for(Int_t i=0;i<TotStp;i++){
00486     if(ff[i]==-1) {
00487       ff[i] = 100.*PredictEMLoss(fMinBoxPlane+i/fNStrips,
00488                                  fMinBoxStrip+i%fNStrips,flucVal);    
00489       fff[i] = 100.*flucVal;
00490     }
00491     for(Int_t j=0;j<TotStp;j++){
00492       if(ff[j]==-1) {
00493         ff[j] = 100.*PredictEMLoss(fMinBoxPlane+j/fNStrips,
00494                                    fMinBoxStrip+j%fNStrips,flucVal);
00495         fff[j] = 100.*flucVal;
00496       }
00497       ErrArray[i*TotStp+j]*=(fff[i]*fff[j]);
00498       if(i==j) {
00499         if(ErrArray[i*TotStp+j]==0) ErrArray[i*TotStp+j] = lastDiagTerm;
00500         else lastDiagTerm = ErrArray[i*TotStp+j];
00501         MSG("FitShowerEM", Msg::kVerbose) << "ErrArray: " 
00502                                           << ErrArray[i*TotStp+j]
00503                                           << endl;      
00504       }
00505       
00506     }
00507   }
00508   
00509   //invert covariance matrix:
00510   Double_t *InvArray = new Double_t[TotStp*TotStp];
00511   TMatrixD matrix(TotStp,TotStp);
00512   matrix.SetMatrixArray(ErrArray);
00513   matrix.Invert();
00514   if(matrix.IsValid()) matrix.GetMatrix2Array(InvArray);
00515   else return -1.0;
00516 
00517   //loop through again to calculate chi2
00518   Double_t Chi2 = 0;
00519   Double_t signStat = 0;
00520   Double_t nValidPoints = 0;
00521   for(Int_t i=0;i<TotStp;i++){
00522     MSG("FitShowerEM", Msg::kVerbose) << "MIP: " << fMips[i] 
00523                                       << " FF: " << ff[i] << endl;
00524     if(fMips[i]>ff[i]) signStat+=1;
00525     nValidPoints+=1;
00526     for(Int_t j=0;j<TotStp;j++){
00527       Chi2+=(fMips[i]-ff[i])*InvArray[i*TotStp+j]*(fMips[j]-ff[j]);
00528     }
00529   }
00530   
00531   fSignStat = signStat/nValidPoints;
00532   fNDF=TotStp;
00533  
00534   delete [] InvArray;
00535   delete [] ff;
00536 
00537   return Chi2;
00538   
00539 }

void FitterEM::DoFit (   ) 
 

Definition at line 542 of file FitterEM.cxx.

References CalculateChi2(), Chi2(), fBestSignStat, fChi2Min, fFittedErr, fFittedPars, fNDF, fNiter, fNSteps, fPass, fStepSize, and MSG.

Referenced by AlgFitShowerEM::RunAlg().

00542                     {
00543 
00544   MSG("FitShowerEM", Msg::kDebug) << "In DoFit()" << endl;
00545 
00546   fFittedErr = new Double_t[NPAR];
00547   
00548   Double_t *BestPars = new Double_t[NPAR];  
00549   Double_t *parlowlim = new Double_t[NPAR];
00550   Double_t *parupplim = new Double_t[NPAR];
00551   Int_t nvarypar = 0;
00552   
00553   for(Int_t i=0;i<NPAR;i++){    
00554     if(fStepSize[i]==0||fNSteps[i]<1) {
00555       parlowlim[i] = fFittedPars[i];
00556       parupplim[i] = fFittedPars[i];
00557     }
00558     else {
00559       parlowlim[i] = fFittedPars[i] - fNSteps[i]*fStepSize[i];
00560       parupplim[i] = fFittedPars[i] + fNSteps[i]*fStepSize[i];      
00561       nvarypar += 1;
00562     }
00563   }
00564 
00565   for(Double_t a=parlowlim[0];a<=parupplim[0];a+=fStepSize[0]){
00566     fFittedPars[0] = a;
00567     for(Double_t b=parlowlim[1];b<=parupplim[1];b+=fStepSize[1]){
00568       fFittedPars[1] = b;
00569       for(Double_t c=parlowlim[2];c<=parupplim[2];c+=fStepSize[2]){
00570         fFittedPars[2] = c;
00571         for(Double_t d=parlowlim[3];d<=parupplim[3];d+=fStepSize[3]){
00572           fFittedPars[3] = d;
00573           for(Double_t e=parlowlim[4];e<=parupplim[4];e+=fStepSize[4]){
00574             fFittedPars[4] = e;
00575             for(Double_t f=parlowlim[5];f<=parupplim[5];f+=fStepSize[5]){
00576               fFittedPars[5] = f;
00577 
00578               MSG("FitShowerEM", Msg::kDebug) 
00579                 <<a<<" "<<b<<" "<<c<<" "<<d<<" "<<e<<" "<<f<<endl;
00580               
00581               if(fFittedPars[0]<0.05) MSG("FitShowerEM", Msg::kWarning) 
00582                 <<"Energy < 50MeV! Not calculating Chi2" << endl;
00583               Double_t Chi2 = CalculateChi2();
00584               fNiter+=1;
00585               if(Chi2==-1) continue; //problem with calculation
00586               if(Chi2<fChi2Min) {
00587                 fChi2Min = Chi2;
00588                 for(Int_t i=0;i<NPAR;i++) BestPars[i] = fFittedPars[i];
00589                 fBestSignStat = fSignStat;
00590                 //DoTests();
00591               }
00592               MSG("FitShowerEM", Msg::kDebug) << "Chi2: "
00593                                               << Chi2 << endl;
00594             }
00595           }
00596         }
00597       }
00598     }
00599   }
00600   
00601   for(Int_t i=0;i<NPAR;i++) fFittedPars[i] = BestPars[i];
00602   fNDF -= nvarypar;
00603 
00604   fPass = true;
00605   if(fChi2Min==99999||fNDF<=0) fPass = false;
00606 
00607 }

void FitterEM::DoTests (   ) 
 

Definition at line 610 of file FitterEM.cxx.

References fKSStat, fMinBoxPlane, fMinBoxStrip, fMips, fNPlanes, fNStrips, fRunStat, and PredictEMLoss().

00610                       {
00611 
00612   Int_t nValidData        = 0;
00613   Int_t nValidFunc        = 0;
00614   Double_t *DataVals      = new Double_t[fNPlanes*fNStrips];
00615   Double_t *DataVals_Asc  = new Double_t[fNPlanes*fNStrips];
00616   Double_t *FuncVals      = new Double_t[fNPlanes*fNStrips];
00617   Double_t *FuncVals_Asc  = new Double_t[fNPlanes*fNStrips];
00618   Double_t lowest_DataVal = 99999;
00619   Double_t lowest_FuncVal = 99999;
00620   Double_t flucVal = 0;
00621 
00622   for(Int_t i=0;i<fNPlanes;i++){
00623     for(Int_t j=1;j<=fNStrips;j++){
00624       Float_t bc = fMips[i*fNStrips+j];
00625       Float_t ff = PredictEMLoss(fMinBoxPlane+i,fMinBoxStrip+j,flucVal);
00626 
00627       if(ff>0){
00628         FuncVals[nValidFunc] = ff;
00629         nValidFunc+=1;
00630         if(ff<lowest_FuncVal) lowest_FuncVal = ff;
00631       }
00632       
00633       if(bc>0){
00634         DataVals[nValidData] = bc;
00635         nValidData+=1;
00636         if(bc<lowest_DataVal) lowest_DataVal = bc;
00637       }
00638     }
00639   }
00640   
00641   //put arrays into ascending order:
00642   DataVals_Asc[0] = lowest_DataVal;
00643   for(Int_t i=1;i<nValidData;i++){
00644     Double_t lowest_DataVal = 999999;
00645     for(Int_t j=0;j<nValidData;j++){
00646       if(DataVals[j]<lowest_DataVal&&DataVals[j]>DataVals_Asc[i-1])
00647         lowest_DataVal = DataVals[j];
00648     }
00649     DataVals_Asc[i] = lowest_DataVal;
00650   }
00651   delete [] DataVals;
00652 
00653   FuncVals_Asc[0] = lowest_FuncVal;
00654   for(Int_t i=1;i<nValidFunc;i++){
00655     Double_t lowest_FuncVal = 999999;
00656     for(Int_t j=0;j<nValidFunc;j++){
00657       if(FuncVals[j]<lowest_FuncVal&&FuncVals[j]>FuncVals_Asc[i-1])
00658         lowest_FuncVal = FuncVals[j];
00659     }
00660     FuncVals_Asc[i] = lowest_FuncVal;
00661   }
00662   delete [] FuncVals;
00663 
00664   //calculate number of runs and mean and variance of expected number of runs:
00665   Double_t mean_runs = 1. + 2.*nValidFunc*nValidData/(nValidData+nValidFunc);
00666   Double_t var_runs = (2.*nValidFunc*nValidData
00667                      *(2.*nValidFunc*nValidData - nValidData -nValidFunc))
00668     /((nValidFunc+nValidData)*(nValidFunc+nValidData)
00669       *(nValidFunc+nValidData-1));
00670 
00671   Double_t r = 0;
00672   Int_t nD = 0;
00673   Int_t nF = 0;
00674   Int_t currentBun = 0; // 1 => Data was last lowest; -1 => Func was last lowest
00675   Bool_t change = false;
00676 
00677   for(Int_t i=0;i<nValidData+nValidFunc;i++){
00678     if(DataVals_Asc[nD]<FuncVals_Asc[nF]) {
00679       nD+=1;
00680       if(currentBun!=1) {
00681         change = true;
00682         currentBun = 1;
00683       }
00684     }
00685     else {
00686       nF+=1;
00687       if(currentBun!=-1) {
00688         change = true;
00689         currentBun = -1;
00690       } 
00691     }
00692     if(change) {
00693       r+=1;
00694       change = false;
00695     }
00696     if(nD>=nValidData) break;
00697     if(nF>=nValidFunc) break;
00698   }
00699 
00700   fRunStat = (r - mean_runs)/sqrt(var_runs);
00701   
00702   //Now do K-S test:
00703   for(Int_t i=0;i<nValidData-1;i++){
00704     //between DataVals_Asc[i] and DataVals_Asc[i+1]
00705     //cumulative distribution value is (i+1)/nValidData
00706     //Estimate from Func by interpolating at 
00707     //(DataVals_Asc[i+1]-DataVals_Asc[i])/2.
00708     Double_t S =  Double_t(i+1)/Double_t(nValidData);
00709     Double_t theVal = (DataVals_Asc[i+1]+DataVals_Asc[i])/2.;
00710     Int_t funcIndex = -1;
00711     for(Int_t j=1;j<nValidFunc;j++){
00712       if(FuncVals_Asc[j]>theVal) {
00713         funcIndex = j;
00714         break;
00715       }
00716     }
00717     if(funcIndex==-1) funcIndex = nValidFunc-1;
00718     Double_t estimator = funcIndex-1. + (FuncVals_Asc[funcIndex]-theVal)
00719       /(FuncVals_Asc[funcIndex]-FuncVals_Asc[funcIndex-1]);
00720     estimator/=Double_t(nValidFunc);
00721     
00722     if(fabs(S-estimator)>fKSStat) fKSStat = fabs(S-estimator);
00723     
00724   }
00725 
00726   delete [] DataVals_Asc;
00727   delete [] FuncVals_Asc;
00728   return;
00729 
00730 }

Double_t FitterEM::GetChi2 (   )  [inline]
 

Definition at line 81 of file FitterEM.h.

Referenced by AlgFitShowerEM::RunAlg().

00081 {return fChi2Min;}

Double_t* FitterEM::GetFittedErr (   )  [inline]
 

Definition at line 84 of file FitterEM.h.

00084 {return fFittedErr;}

Double_t * FitterEM::GetFittedPars (   ) 
 

Definition at line 244 of file FitterEM.cxx.

References fFittedPars, and fVtx.

Referenced by AlgFitShowerEM::RunAlg().

00244                                  {
00245   //convert back from internal coords to CandFitShowerEM coords
00246   fFittedPars[0] = fFittedPars[0]; //energy is ok
00247   fFittedPars[1] = fFittedPars[1]*0.041+fVtx[0]-fclosestTPosUDiff;//u vertex
00248   fFittedPars[2] = fFittedPars[2]*0.041+fVtx[1]-fclosestTPosVDiff;//v vertex
00249   fFittedPars[3] = fFittedPars[3]*0.060+fVtx[2]-fclosestZedDiff;//z vertex
00250   fFittedPars[4] = TMath::Tan(fFittedPars[4])*4.1/6.0; //dudz
00251   fFittedPars[5] = TMath::Tan(fFittedPars[5])*4.1/6.0;//dvdz
00252   return fFittedPars; 
00253 }

Double_t FitterEM::GetKSTestStat (   )  [inline]
 

Definition at line 90 of file FitterEM.h.

00090 {return fKSStat;}

Int_t FitterEM::GetNDF (   )  [inline]
 

Definition at line 82 of file FitterEM.h.

Referenced by AlgFitShowerEM::RunAlg().

00082 {return fNDF;}

Int_t FitterEM::GetNIterations (   )  [inline]
 

Definition at line 85 of file FitterEM.h.

Referenced by AlgFitShowerEM::RunAlg().

00085 {return fNiter;}

Double_t FitterEM::GetRunTestStat (   )  [inline]
 

Definition at line 89 of file FitterEM.h.

00089 {return fRunStat;}

Double_t FitterEM::GetSignTestStat (   )  [inline]
 

Definition at line 88 of file FitterEM.h.

00088 {return fBestSignStat;}

Bool_t FitterEM::Pass (   )  [inline]
 

Definition at line 80 of file FitterEM.h.

Referenced by AlgFitShowerEM::RunAlg().

00080 {return fPass;}

Double_t FitterEM::PredictEMLoss (  Int_t  ,
Int_t  ,
Double_t & 
) 
 

Definition at line 425 of file FitterEM.cxx.

References CacherEM::ClearCache(), CacherEM::ClearPSCache(), CUTOFF, DELTA_PHI, DELTA_R, DELTA_T, EPSILON, fCache, fclosestZedView, fFittedPars, CacherEM::GetPSMap(), CacherEM::GetRTMap(), CacherEM::GetSumE(), MakeKey(), NPAR, CacherEM::PrintPSMap(), CacherEM::PrintRTMap(), PSCalc(), RTCalc(), CacherEM::StorePS(), CacherEM::StoreRT(), CacherEM::ValidPS(), and CacherEM::ValidRT().

Referenced by CalculateChi2(), DoTests(), MadHandScan::HandScanDisplay(), and NueDisplayModule::IntRecoCalc().

00426 {
00427 
00428   if(!fCache->ValidPS(fFittedPars)){
00429     fCache->ClearPSCache();
00430     double sumE = 0;
00431     vector<rte> rtmap;
00432     
00433     if(!fCache->ValidRT(fFittedPars)){
00434       fCache->ClearCache();
00435       rtmap = RTCalc(fFittedPars[0],DELTA_T,DELTA_R,EPSILON,CUTOFF,sumE);
00436       fCache->StoreRT(NPAR,fFittedPars,rtmap,sumE);
00437       fCache->PrintRTMap();
00438     }
00439     else {
00440       rtmap = fCache->GetRTMap();
00441       sumE = fCache->GetSumE();
00442     }
00443     
00444     map<int,spef> psmap = PSCalc(rtmap,fclosestZedView,
00445                                  fFittedPars[1],fFittedPars[2],
00446                                  fFittedPars[3],fFittedPars[4],
00447                                  fFittedPars[5],DELTA_PHI,sumE);
00448     fCache->StorePS(NPAR,fFittedPars,psmap);
00449     fCache->PrintPSMap();
00450   }
00451  
00452   map<int,spef> spmap = fCache->GetPSMap();
00453   int key = MakeKey(plane,strip);
00454   map<int,spef>::iterator theOne = spmap.find(key);
00455   map<int,spef>::iterator end = spmap.end();
00456   if(theOne==end) {
00457     flucVal = 0.;
00458     return 0;
00459   }
00460   flucVal = theOne->second.fl;
00461   return theOne->second.en;
00462   
00463 }

void FitterEM::QuickInput (  Double_t  ,
Double_t  ,
Double_t  ,
Double_t  ,
Double_t  ,
Double_t  ,
Int_t 
) 
 

Definition at line 218 of file FitterEM.cxx.

References fclosestZedView, fFittedPars, and MSG.

Referenced by MadHandScan::HandScanDisplay(), and NueDisplayModule::IntRecoCalc().

00220                                       {
00221   
00222   fFittedPars = new Double_t[NPAR];
00223   fFittedPars[0] = energy;
00224   fFittedPars[1] = vtxu/0.041; //vertex
00225   fFittedPars[2] = vtxv/0.041;
00226   fFittedPars[3] = vtxz/0.060;
00227   fFittedPars[4] = TMath::ATan(dudz*6.0/4.1); //angles in rads 
00228   fFittedPars[5] = TMath::ATan(dvdz*6.0/4.1); //in pln-stp coords
00229   
00230   fclosestZedView = view;
00231 
00232   MSG("FitShowerEM", Msg::kDebug) 
00233     << "Fit Parameters to start with:\n" 
00234     << "Energy = "<<fFittedPars[0]<<"\n"
00235     << "UVtx   = "<<fFittedPars[1]<<"\n"
00236     << "VVtx   = "<<fFittedPars[2]<<"\n"
00237     << "ZVtx   = "<<fFittedPars[3]<<"\n"
00238     << "UAng   = "<<fFittedPars[4]<<"\n"
00239     << "VAng   = "<<fFittedPars[5]<<endl;
00240 
00241 }

void FitterEM::SetInputParams (  Double_t  ,
Double_t *  ,
Double_t  ,
Double_t 
) 
 

Definition at line 205 of file FitterEM.cxx.

References fdudz, fdvdz, fEnergy, fVtx, and SetUpFitBox().

Referenced by AlgFitShowerEM::RunAlg().

00206                                                            {
00207   fEnergy = energy;
00208   fVtx = new Double_t[3]; 
00209   fVtx[0] = vtx[0]; 
00210   fVtx[1] = vtx[1]; 
00211   fVtx[2] = vtx[2];
00212   fdudz = dudz; 
00213   fdvdz = dvdz;
00214   SetUpFitBox();
00215 }

void FitterEM::SetNSteps (  Int_t *   ) 
 

Definition at line 278 of file FitterEM.cxx.

References fNSteps, and MSG.

Referenced by AlgFitShowerEM::RunAlg().

00278                                       {
00279   
00280   fNSteps = new Int_t[NPAR];
00281   for(int i=0;i<NPAR;i++) fNSteps[i] = NSteps[i];
00282 
00283   MSG("FitShowerEM",Msg::kDebug) << "Number of Steps: " << fNSteps[0] 
00284                                  << " " << fNSteps[1] << " " << fNSteps[2]
00285                                  << " " << fNSteps[3] << " " << fNSteps[4]
00286                                  << " " << fNSteps[5] << endl;
00287 }

void FitterEM::SetStepSize (  Double_t *   ) 
 

Definition at line 256 of file FitterEM.cxx.

References fStepSize, and MSG.

Referenced by AlgFitShowerEM::RunAlg().

00256                                              {
00257   
00258   fStepSize = new Double_t[NPAR];
00259   for(int i=0;i<NPAR;i++) fStepSize[i] = stepSize[i];
00260   
00261   fStepSize[0] *= fEnergy;           //assume a %    --> GeV
00262   fStepSize[1] /= 0.041;             //assume in m   --> units of strip
00263   fStepSize[2] /= 0.041;             //assume in m   --> units of strip
00264   fStepSize[3] /= 0.060;             //assume in m   --> units of plane
00265   
00266   //assume in deg in metre-metre --> rads in plane-strip
00267   fStepSize[4]=TMath::ATan(TMath::Tan(fStepSize[4]*TMath::Pi()/180.)*6.0/4.1);
00268   fStepSize[5]=TMath::ATan(TMath::Tan(fStepSize[5]*TMath::Pi()/180.)*6.0/4.1);
00269   
00270   MSG("FitShowerEM",Msg::kDebug) << "Step Sizes: " << fStepSize[0] 
00271                                  << " " << fStepSize[1] << " " << fStepSize[2]
00272                                  << " " << fStepSize[3] << " " << fStepSize[4]
00273                                  << " " << fStepSize[5] << endl;
00274   
00275 }

void FitterEM::SetUpFitBox (   )  [private]
 

Definition at line 290 of file FitterEM.cxx.

References abs(), fBinCorrel, fBoxPlaneArray, fBoxStripArray, fclosestPlaneToVtx, fclosestTPosUDiff, fclosestTPosVDiff, fclosestUStripToVtx, fclosestVStripToVtx, fclosestZedDiff, fclosestZedView, fdudz, fdvdz, fFittedPars, fMaxBoxPlane, fMaxBoxStrip, fMinBoxPlane, fMinBoxStrip, fMipArray, fMips, fNPlanes, fNStrips, fPlaneArray, fStripArray, fTPosArray, fViewArray, fVtx, fZedArray, and MSG.

Referenced by SetInputParams().

00290                            {
00291 
00292   //This is the tricky part...
00293   //need to move into (plane,strip) Box with vertex approx at (0,0,0)
00294   //adjust all strip numbers by appropriate transverse vtx
00295    //adjust all plane numbers by z vtx
00296   //adjust angles to be u-z, v-z projection angles in rads 
00297   //calculated in units of plane and strip
00298   //need to set fFittedPars to be the new angles and vtx in the Box coords
00299   
00300   for(int i=0;i<fNStp;i++){
00301     float ZedDiff = fZedArray[i]-fVtx[2];
00302     if(fabs(ZedDiff)<fabs(fclosestZedDiff)){
00303       fclosestZedDiff = ZedDiff;
00304       fclosestPlaneToVtx = fPlaneArray[i];
00305       fclosestZedView = fViewArray[i];
00306     }
00307   }
00308   
00309   Int_t closestOtherViewPlaneToVtxPlane = 99999;
00310   for(int i=0;i<fNStp;i++){
00311     //if this is the vertex plane, use it for finding transverse vertex
00312     if(fPlaneArray[i]==fclosestPlaneToVtx){
00313       if(fViewArray[i]==2){ //U strips
00314         float TPosDiff = fTPosArray[i]-fVtx[0];
00315         if(fabs(TPosDiff)<fabs(fclosestTPosUDiff)){
00316           fclosestTPosUDiff = TPosDiff;
00317           fclosestUStripToVtx = fStripArray[i];
00318         }
00319       }
00320       else if(fViewArray[i]==3){ //V strips
00321         float TPosDiff = fTPosArray[i]-fVtx[1];
00322         if(fabs(TPosDiff)<fabs(fclosestTPosVDiff)){
00323           fclosestTPosVDiff = TPosDiff;
00324           fclosestVStripToVtx = fStripArray[i];
00325         }
00326       }
00327     }
00328     //otherwise, if plane-view is not same as vertex plane-view, 
00329     //then use it to find vertex in other view:
00330     else if(fViewArray[i]!=fclosestZedView){
00331       //look for closest plane in other view:
00332       if(abs(fPlaneArray[i]-fclosestPlaneToVtx)<= 
00333          abs(closestOtherViewPlaneToVtxPlane-fclosestPlaneToVtx)){
00334         closestOtherViewPlaneToVtxPlane=fPlaneArray[i];
00335         if(fViewArray[i]==2){ //U strips
00336           float TPosDiff = fTPosArray[i]-fVtx[0];
00337           if(fabs(TPosDiff)<fabs(fclosestTPosUDiff)){
00338             fclosestTPosUDiff = TPosDiff;
00339             fclosestUStripToVtx = fStripArray[i];
00340           }
00341         }
00342         else if(fViewArray[i]==3){ //V strips
00343           float TPosDiff = fTPosArray[i]-fVtx[1];
00344           if(fabs(TPosDiff)<fabs(fclosestTPosVDiff)){
00345             fclosestTPosVDiff = TPosDiff;
00346             fclosestVStripToVtx = fStripArray[i];
00347           }
00348         }
00349       }
00350     }
00351   }
00352   
00353 
00354   MSG("FitShowerEM", Msg::kDebug) 
00355     << "Box Quantities: Closest Plane To Vtx="<<fclosestPlaneToVtx
00356     << " with ZedDiff="<<fclosestZedDiff
00357     << " and in View="<<fclosestZedView<<"\n"
00358     << "Closest U Strip to Vtx="<<fclosestUStripToVtx
00359     << " with TPosDiff="<<fclosestTPosUDiff<<"\n"
00360     << "Closest V Strip to Vtx="<<fclosestVStripToVtx
00361     << " with TPosDiff="<<fclosestTPosVDiff<<"\n"
00362     << "(Vertex from CandShowerEM=["<<fVtx[0]<<","<<fVtx[1]<<","<<fVtx[2]<<"])"
00363     << endl;
00364   
00365   for(int i=0;i<fNStp;i++){
00366     
00367     fBoxPlaneArray[i] = fPlaneArray[i]-fclosestPlaneToVtx;
00368     if(fViewArray[i]==2) {//U strips  
00369       fBoxStripArray[i] = fStripArray[i]-fclosestUStripToVtx; 
00370     }
00371     if(fViewArray[i]==3) {//V strips
00372       fBoxStripArray[i] = fStripArray[i]-fclosestVStripToVtx; 
00373     }
00374     
00375     if(fBoxPlaneArray[i]<fMinBoxPlane) fMinBoxPlane = fBoxPlaneArray[i];
00376     if(fBoxPlaneArray[i]>fMaxBoxPlane) fMaxBoxPlane = fBoxPlaneArray[i];
00377     if(fBoxStripArray[i]<fMinBoxStrip) fMinBoxStrip = fBoxStripArray[i];
00378     if(fBoxStripArray[i]>fMaxBoxStrip) fMaxBoxStrip = fBoxStripArray[i];
00379   }
00380 
00381 
00382   fNPlanes = fMaxBoxPlane-fMinBoxPlane+1;
00383   fNStrips = fMaxBoxStrip-fMinBoxStrip+1;
00384   
00385   //now make fMips[plane*strip] array to simplify Mip value access in fit
00386   Int_t TotStp = fNPlanes*fNStrips;
00387   fMips = new Double_t[TotStp];
00388   for(int i=0;i<TotStp;i++) fMips[i] = 0;
00389   for(int i=0;i<fNStp;i++){    
00390     int n = ((fBoxPlaneArray[i]-fMinBoxPlane)*fNStrips 
00391              + (fBoxStripArray[i] - fMinBoxStrip));
00392     fMips[n] = 100.*fMipArray[i]/fTotMip; //make everything a percentage
00393   }
00394   
00395   MSG("FitShowerEM", Msg::kDebug) 
00396     << "Box Limits: "<<"\n"
00397     << "NPlanes="<<fNPlanes<<" with Min="<<fMinBoxPlane
00398     << " and Max="<<fMaxBoxPlane<<"\n"
00399     << "NStrips="<<fNStrips<<" with Min="<<fMinBoxStrip
00400     << " and Max="<<fMaxBoxStrip<<endl;
00401 
00402   fFittedPars = new Double_t[NPAR];
00403   fFittedPars[0] = fEnergy;
00404   fFittedPars[1] = fclosestTPosUDiff/0.041; //vertex
00405   fFittedPars[2] = fclosestTPosVDiff/0.041;
00406   fFittedPars[3] = fclosestZedDiff/0.060;  
00407   fFittedPars[4] = TMath::ATan(fdudz*6.0/4.1); //angles in rads 
00408   fFittedPars[5] = TMath::ATan(fdvdz*6.0/4.1); //in pln-stp coords
00409   
00410   MSG("FitShowerEM", Msg::kDebug) 
00411     << "Fit Parameters to start with:\n" 
00412     << "Energy = "<<fFittedPars[0]<<"\n"
00413     << "UVtx   = "<<fFittedPars[1]<<"\n"
00414     << "VVtx   = "<<fFittedPars[2]<<"\n"
00415     << "ZVtx   = "<<fFittedPars[3]<<"\n"
00416     << "UAng   = "<<fFittedPars[4]<<"\n"
00417     << "VAng   = "<<fFittedPars[5]<<endl;
00418 
00419   fBinCorrel = new BinCorrelationEM(fFittedPars);
00420 
00421 }

void FitterEM::SetUpStripInfo (   )  [private]
 

Definition at line 168 of file FitterEM.cxx.

References fBoxPlaneArray, fBoxStripArray, fDataStpItr, fMipArray, fPlaneArray, fStripArray, fTotMip, fTPosArray, fViewArray, fZedArray, CandStripHandle::GetCharge(), Calibrator::GetMIP(), CandStripHandle::GetPlane(), CandStripHandle::GetPlaneView(), CandStripHandle::GetStrip(), CandStripHandle::GetTPos(), CandHandle::GetVldContext(), CandStripHandle::GetZPos(), Calibrator::Instance(), MSG, and CalScheme::Reset().

Referenced by FitterEM().

00168                              {
00169 
00170   Int_t icnt = 0;
00171   Calibrator& cal=Calibrator::Instance();
00172   
00173   while (CandStripHandle *stp = fDataStpItr()) {
00174     
00175     if(icnt==0) {
00176       const VldContext *vc = stp->GetVldContext();
00177       cal.Reset(*vc);
00178     }
00179 
00180     int plane = stp->GetPlane();
00181     int strip = stp->GetStrip();
00182     fPlaneArray[icnt] = plane;
00183     fStripArray[icnt] = strip;
00184     fViewArray[icnt]  = stp->GetPlaneView(); //U == 2, V == 3;
00185     fMipArray[icnt]   = cal.GetMIP(stp->GetCharge(CalDigitType::kSigCorr,
00186                                                   StripEnd::kWhole));
00187     fTotMip += fMipArray[icnt];
00188     fZedArray[icnt]   = stp->GetZPos();
00189     fTPosArray[icnt]  = stp->GetTPos();
00190     fBoxPlaneArray[icnt] = 0;
00191     fBoxStripArray[icnt] = 0;
00192 
00193     MSG("FitShowerEM", Msg::kVerbose) << "pl="<<fPlaneArray[icnt]
00194                                       <<" st="<<fStripArray[icnt]
00195                                       <<" pv="<<fViewArray[icnt]
00196                                       <<" mips="<<fMipArray[icnt]
00197                                       <<endl;
00198     
00199     icnt++;
00200   }
00201 
00202 }

Member Data Documentation

Double_t FitterEM::fBestSignStat [private]
 

Definition at line 57 of file FitterEM.h.

Referenced by DoFit().

BinCorrelationEM* FitterEM::fBinCorrel [private]
 

Definition at line 13 of file FitterEM.h.

Referenced by CalculateChi2(), FitterEM(), and SetUpFitBox().

Int_t* FitterEM::fBoxPlaneArray [private]
 

Definition at line 43 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Int_t* FitterEM::fBoxStripArray [private]
 

Definition at line 44 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

CacherEM* FitterEM::fCache [private]
 

Definition at line 12 of file FitterEM.h.

Referenced by FitterEM(), and PredictEMLoss().

Double_t FitterEM::fChi2Min [private]
 

Definition at line 52 of file FitterEM.h.

Referenced by DoFit(), and FitterEM().

Int_t FitterEM::fclosestPlaneToVtx [private]
 

Definition at line 21 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Float_t FitterEM::fclosestTPosUDiff [private]
 

Definition at line 26 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Float_t FitterEM::fclosestTPosVDiff [private]
 

Definition at line 27 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Int_t FitterEM::fclosestUStripToVtx [private]
 

Definition at line 24 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Int_t FitterEM::fclosestVStripToVtx [private]
 

Definition at line 25 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Float_t FitterEM::fclosestZedDiff [private]
 

Definition at line 22 of file FitterEM.h.

Referenced by FitterEM(), and SetUpFitBox().

Int_t FitterEM::fclosestZedView [private]
 

Definition at line 23 of file FitterEM.h.

Referenced by FitterEM(), PredictEMLoss(), QuickInput(), and SetUpFitBox().

CandStripHandleItr FitterEM::fDataStpItr [private]
 

Definition at line 11 of file FitterEM.h.

Referenced by FitterEM(), and SetUpStripInfo().

Double_t FitterEM::fdudz [private]
 

Definition at line 18 of file FitterEM.h.

Referenced by FitterEM(), SetInputParams(), and SetUpFitBox().

Double_t FitterEM::fdvdz [private]
 

Definition at line 19 of file FitterEM.h.

Referenced by FitterEM(), SetInputParams(), and SetUpFitBox().

Double_t FitterEM::fEnergy [private]
 

Definition at line 16 of file FitterEM.h.

Referenced by FitterEM(), and SetInputParams().

Double_t* FitterEM::fFittedErr [private]
 

Definition at line 49 of file FitterEM.h.

Referenced by DoFit(), and FitterEM().

Double_t* FitterEM::fFittedPars [private]
 

Definition at line 48 of file FitterEM.h.

Referenced by CalculateChi2(), DoFit(), FitterEM(), GetFittedPars(), PredictEMLoss(), QuickInput(), and SetUpFitBox().

Double_t FitterEM::fKSStat [private]
 

Definition at line 59 of file FitterEM.h.

Referenced by DoTests(), and FitterEM().

Int_t FitterEM::fMaxBoxPlane [private]
 

Definition at line 32 of file FitterEM.h.

Referenced by CalculateChi2(), FitterEM(), and SetUpFitBox().

Int_t FitterEM::fMaxBoxStrip [private]
 

Definition at line 34 of file FitterEM.h.

Referenced by CalculateChi2(), FitterEM(), and SetUpFitBox().

Int_t FitterEM::fMinBoxPlane [private]
 

Definition at line 31 of file FitterEM.h.

Referenced by CalculateChi2(), DoTests(), FitterEM(), and SetUpFitBox().

Int_t FitterEM::fMinBoxStrip [private]
 

Definition at line 33 of file FitterEM.h.

Referenced by CalculateChi2(), DoTests(), FitterEM(), and SetUpFitBox().

Double_t* FitterEM::fMipArray [private]
 

Definition at line 40 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Double_t* FitterEM::fMips [private]
 

Definition at line 45 of file FitterEM.h.

Referenced by CalculateChi2(), DoTests(), FitterEM(), and SetUpFitBox().

Int_t FitterEM::fNDF [private]
 

Definition at line 53 of file FitterEM.h.

Referenced by CalculateChi2(), DoFit(), and FitterEM().

Int_t FitterEM::fNiter [private]
 

Definition at line 54 of file FitterEM.h.

Referenced by DoFit(), and FitterEM().

Int_t FitterEM::fNPlanes [private]
 

Definition at line 29 of file FitterEM.h.

Referenced by CalculateChi2(), DoTests(), FitterEM(), and SetUpFitBox().

Int_t* FitterEM::fNSteps [private]
 

Definition at line 51 of file FitterEM.h.

Referenced by DoFit(), FitterEM(), and SetNSteps().

Int_t FitterEM::fNStp [private]
 

Definition at line 15 of file FitterEM.h.

Referenced by FitterEM().

Int_t FitterEM::fNStrips [private]
 

Definition at line 30 of file FitterEM.h.

Referenced by DoTests(), FitterEM(), and SetUpFitBox().

Bool_t FitterEM::fPass [private]
 

Definition at line 47 of file FitterEM.h.

Referenced by DoFit(), and FitterEM().

Int_t* FitterEM::fPlaneArray [private]
 

Definition at line 37 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Double_t FitterEM::fRunStat [private]
 

Definition at line 58 of file FitterEM.h.

Referenced by DoTests(), and FitterEM().

Double_t FitterEM::fSignStat [private]
 

Definition at line 56 of file FitterEM.h.

Referenced by CalculateChi2(), and FitterEM().

Double_t* FitterEM::fStepSize [private]
 

Definition at line 50 of file FitterEM.h.

Referenced by DoFit(), FitterEM(), and SetStepSize().

Int_t* FitterEM::fStripArray [private]
 

Definition at line 38 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Double_t FitterEM::fTotMip [private]
 

Definition at line 36 of file FitterEM.h.

Referenced by FitterEM(), and SetUpStripInfo().

Double_t* FitterEM::fTPosArray [private]
 

Definition at line 42 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Int_t* FitterEM::fViewArray [private]
 

Definition at line 39 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

Double_t* FitterEM::fVtx [private]
 

Definition at line 17 of file FitterEM.h.

Referenced by FitterEM(), GetFittedPars(), SetInputParams(), and SetUpFitBox().

Double_t* FitterEM::fZedArray [private]
 

Definition at line 41 of file FitterEM.h.

Referenced by FitterEM(), SetUpFitBox(), and SetUpStripInfo().

The documentation for this class was generated from the following files:
Generated on Mon Feb 15 11:09:13 2010 for loon by  doxygen 1.3.9.1