2 // **************************************************************************
3 // * Copyright(c) 1998-1999, ALICE Experiment at CERN, All rights reserved. *
5 // * Author: The ALICE Off-line Project. *
6 // * Contributors are mentioned in the code where appropriate. *
8 // * Permission to use, copy, modify and distribute this software and its *
9 // * documentation strictly for non-commercial purposes is hereby granted *
10 // * without fee, provided that the above copyright notice appears in all *
11 // * copies and that both the copyright notice and this permission notice *
12 // * appear in the supporting documentation. The authors make no claims *
13 // * about the suitability of this software for any purpose. It is *
14 // * provided "as is" without express or implied warranty. *
15 // **************************************************************************
17 //-----------------------------------------------------------------------------------
18 // Jet finder based on Deterministic Annealing
19 // For further informations about the DA working features see:
20 // Phys.Lett. B601 (2004) 56-63 (http://arxiv.org/abs/hep-ph/0407214)
21 // Author: Davide Perrino (davide.perrino@ba.infn.it, davide.perrino@cern.ch)
22 //-----------------------------------------------------------------------------------
26 #include <TClonesArray.h>
27 #include "AliJetReaderHeader.h"
28 #include "AliJetReader.h"
29 #include "AliDAJetHeader.h"
30 #include "AliDAJetFinder.h"
32 ClassImp(AliDAJetFinder)
35 //-----------------------------------------------------------------------------------
36 AliDAJetFinder::AliDAJetFinder():
52 //-----------------------------------------------------------------------------------
53 AliDAJetFinder::~AliDAJetFinder()
58 //-----------------------------------------------------------------------------------
59 void AliDAJetFinder::FindJets()
61 // Find the jets in current event
63 Float_t betaStop=100.;
64 fDebug = fHeader->GetDebug();
68 TVectorD *vPx = new TVectorD();
69 TVectorD *vPy = new TVectorD();
70 TMatrixD *mPyx= new TMatrixD();
71 TMatrixD *mY = new TMatrixD();
72 InitDetAnn(dEtSum,xData,vPx,vPy,mPyx,mY);
73 if (fNin < fNclustMax){
74 delete [] xData[0], delete [] xData[1];
82 DoubleClusters(nc,nk,vPy,mY);
85 Annealing(nk,xData,vPx,vPy,mPyx,mY);
86 NumCl(nc,nk,vPy,mPyx,mY);
87 }while((fBeta<betaStop || nc<4) && nc<fNclustMax);
89 Int_t *xx=new Int_t[fNeff];
90 for (Int_t i = 0; i < fNeff; i++) xx[i] = 0;
92 EndDetAnn(nk,xData,xx,dEtSum,vPx,vPy,mPyx,mY);
93 StoreJets(nk,xData,xx,mY);
96 delete [] xData[0], delete [] xData[1];
104 //-----------------------------------------------------------------------------------
105 void AliDAJetFinder::InitDetAnn(Double_t &dEtSum,Double_t **xData,TVectorD *vPx,TVectorD *vPy,TMatrixD *mPyx,TMatrixD *mY)
107 //Initialise the variables used by the algorithm
109 fNclustMax = ((AliDAJetHeader*)fHeader)->GetFixedCl() ?
110 ((AliDAJetHeader*)fHeader)->GetNclustMax() :
111 TMath::Max((Int_t)TMath::Sqrt(fNin),5);
112 Float_t etaEff = ((AliDAJetHeader*)fHeader)->GetEtaEff();
113 TClonesArray *lvArray = fReader->GetMomentumArray();
114 Int_t nEntr = lvArray->GetEntries();
116 for (Int_t iEn=0; iEn<nEntr; iEn++) if (fReader->GetCutFlag(iEn)==1) fNin++;
118 fNeff = ((AliDAJetHeader*)fHeader)->GetNeff();
119 fNeff = TMath::Max(fNeff,fNin);
120 Double_t *xEta = new Double_t[fNeff];
121 Double_t *xPhi = new Double_t[fNeff];
122 xData[0]=xEta; xData[1]=xPhi;
123 vPx->ResizeTo(fNeff);
125 for (Int_t iEn=0; iEn<nEntr; iEn++){
126 if (fReader->GetCutFlag(iEn)==0) continue;
127 TLorentzVector *lv=(TLorentzVector*)lvArray->At(iEn);
128 xEta[iIn] = lv->Eta();
129 xPhi[iIn] = lv->Phi()<0 ? lv->Phi() + 2*TMath::Pi() : lv->Phi();
130 (*vPx)(iIn)=lv->Pt();
136 for (iIn=fNin; iIn<fNeff; iIn++){
137 xEta[iIn]=r.Uniform(-1*etaEff,etaEff);
138 xPhi[iIn]=r.Uniform(0.,2*TMath::Pi());
139 (*vPx)(iIn)=r.Uniform(0.01,0.02);
142 for (iIn=0; iIn<fNeff; iIn++) (*vPx)(iIn)=(*vPx)(iIn)/dEtSum;
144 Int_t njdim=2*fNclustMax+1;
145 mPyx->ResizeTo(fNeff,njdim);
146 mY->ResizeTo(4,njdim);
147 vPy->ResizeTo(njdim);
148 mY->Zero();mPyx->Zero();vPy->Zero();
150 TMatrixDColumn(*mPyx,0)=1;
151 Double_t ypos=0,xpos=0;
152 for (iIn=0; iIn<fNeff; iIn++){
153 (*mY)(0,0)+=(*vPx)(iIn)*xEta[iIn];
154 ypos+=(*vPx)(iIn)*TMath::Sin(xPhi[iIn]);
155 xpos+=(*vPx)(iIn)*TMath::Cos(xPhi[iIn]);
157 (*mY)(1,0)=(atan2(ypos,xpos)>0) ? atan2(ypos,xpos) : atan2(ypos,xpos)+2*TMath::Pi();
160 //-----------------------------------------------------------------------------------
161 void AliDAJetFinder::DoubleClusters(Int_t nc,Int_t &nk, TVectorD *vPy, TMatrixD *mY) const
163 // Return double clusters
164 for(Int_t iClust=0; iClust<nc; iClust++){
165 (*vPy)(iClust)=(*vPy)(iClust)/2;
166 (*vPy)(nc+iClust)=(*vPy)(iClust);
167 for(Int_t iComp=0; iComp<3; iComp++) (*mY)(iComp,nc+iClust)=(*mY)(iComp,iClust);
172 //-----------------------------------------------------------------------------------
173 void AliDAJetFinder::Annealing(Int_t nk,Double_t **xData, TVectorD *vPx, TVectorD *vPy, TMatrixD *mPyx, TMatrixD *mY)
175 // Main part of the algorithm
176 const Double_t pi=TMath::Pi();
177 TVectorD *py = new TVectorD(nk);
178 TVectorD *p = new TVectorD(nk);
179 TMatrixD *y = new TMatrixD(4,nk);
180 TMatrixD *y1 = new TMatrixD(4,nk);
181 TMatrixD *ry = new TMatrixD(2,nk);
182 Double_t *xEta = xData[0];
183 Double_t *xPhi = xData[1];
184 Double_t Dist(TVectorD,TVectorD);
186 Double_t df[2]={fReader->GetReaderHeader()->GetFiducialEtaMax(),pi};
188 Double_t *m = new Double_t[nk];
192 for (Int_t iClust=0; iClust<nk; iClust++){
193 for (Int_t i=0; i<3; i++)(*y1)(i,iClust)=(*mY)(i,iClust);
194 (*py)(iClust)=(*vPy)(iClust);
196 //perturbation of codevectors
197 Double_t seed=1000000*gRandom->Rndm(24);
198 ry->Randomize(-0.5,0.5,seed);
199 for (Int_t i=0; i<2; i++){
200 for (Int_t iClust=0; iClust<nk/2; iClust++)
201 (*y1)(i,iClust)+=((*ry)(i,iClust)+TMath::Sign(0.5,(*ry)(i,iClust)))*fDelta*df[i];
202 for (Int_t iClust=nk/2; iClust<nk; iClust++)
203 (*y1)(i,iClust)-=((*ry)(i,iClust-nk/2)+TMath::Sign(0.5,(*ry)(i,iClust-nk/2)))*fDelta*df[i];
206 //recalculate conditional probabilities
208 for (Int_t iIn=0; iIn<fNeff; iIn++){
209 vPart(0)=xEta[iIn]; vPart(1)=xPhi[iIn];
210 for(Int_t iClust=0; iClust<nk; iClust++){
211 (*mPyx)(iIn,iClust)=-log((*py)(iClust))+fBeta*Dist(vPart,TMatrixDColumn(*y1,iClust));
212 m[iClust]=(*mPyx)(iIn,iClust);
215 Double_t minPyx=TMath::MinElement(nk,m);
216 for (Int_t iClust=0; iClust<nk; iClust++){
217 (*mPyx)(iIn,iClust)-=minPyx;
218 (*mPyx)(iIn,iClust)=exp(-(*mPyx)(iIn,iClust));
219 pyxNorm+=(*mPyx)(iIn,iClust);
221 for (Int_t iClust=0; iClust<nk; iClust++) (*mPyx)(iIn,iClust)/=pyxNorm;
225 //recalculate codevectors
226 for (Int_t iClust=0; iClust<nk; iClust++){
227 Double_t xpos=0,ypos=0,pxy;
228 for (Int_t iIn=0; iIn<fNeff; iIn++) (*p)(iClust)+=(*vPx)(iIn)*(*mPyx)(iIn,iClust);
229 for (Int_t iIn=0; iIn<fNeff; iIn++){
230 pxy=(*vPx)(iIn)*(*mPyx)(iIn,iClust)/(*p)(iClust);
231 ypos+=pxy*TMath::Sin(xPhi[iIn]);
232 xpos+=pxy*TMath::Cos(xPhi[iIn]);
233 (*y)(0,iClust)+=pxy*xEta[iIn];
235 (*y)(1,iClust)=(atan2(ypos,xpos)>0) ? atan2(ypos,xpos) : atan2(ypos,xpos)+2*pi;
237 //verify codevectors' stability
239 for (Int_t iClust=0; iClust<nk; iClust++){
240 chi1=TMath::CosH((*y1)(0,iClust)-(*y)(0,iClust))-TMath::Cos((*y1)(1,iClust)-(*y)(1,iClust));
241 chi1/=(2*TMath::CosH((*y1)(0,iClust))*TMath::CosH((*y)(0,iClust)));
243 if (chi1>chi) chi=chi1;
245 chi=TMath::Sqrt(chi);
246 for (Int_t iClust=0; iClust<nk; iClust++){
247 for (Int_t i=0; i<2; i++) (*y1)(i,iClust)=(*y)(i,iClust);
248 (*py)(iClust)=(*p)(iClust);
250 if (nloop>fNloopMax){
251 if (chi<fEpsMax || nloop>500) break;
254 }while (chi>fEpsMax);
255 for (Int_t iClust=0; iClust<nk; iClust++){ //set codevectors and probability equal to those calculated
256 for (Int_t i=0; i<2; i++) (*mY)(i,iClust)=(*y)(i,iClust);
257 (*vPy)(iClust)=(*p)(iClust);
267 //-----------------------------------------------------------------------------------
268 void AliDAJetFinder::NumCl(Int_t &nc,Int_t &nk,TVectorD *vPy, TMatrixD *mPyx,TMatrixD *mY)
270 // Number of clusters
271 static Bool_t growcl=true;
273 if (nk==2) growcl=true;
275 //verify if two codevectors are equal within fAvDist
276 Int_t *nSame = new Int_t[nk];
277 Int_t **iSame = new Int_t*[nk];
278 Int_t **cont = new Int_t*[nk];
279 for (Int_t iClust=0; iClust<nk; iClust++) {
280 cont[iClust] =new Int_t[nk];
281 iSame[iClust]=new Int_t[nk];
284 for (Int_t iClust=0; iClust<nk; iClust++){
285 iSame[iClust][iClust]=1;
286 for (Int_t iClust1=iClust+1; iClust1<nk; iClust1++){
287 Double_t eta = (*mY)(0,iClust) ; Double_t phi = (*mY)(1,iClust);
288 Double_t eta1 = (*mY)(0,iClust1); Double_t phi1 = (*mY)(1,iClust1);
289 Double_t distCl=(TMath::CosH(eta-eta1)-TMath::Cos(phi-phi1))/(2*TMath::CosH(eta)*TMath::CosH(eta1));
290 if (distCl < fAvDist) iSame[iClust][iClust1]=iSame[iClust1][iClust]=1;
291 else iSame[iClust][iClust1]=iSame[iClust1][iClust]=0;
294 ReduceClusters(iSame,nk,nc,cont,nSame);
295 if (nc >= fNclustMax) growcl=false;
296 //recalculate the nc distinct codevectors
297 TMatrixD *pyx = new TMatrixD(fNeff,2*nc);
298 TVectorD *py = new TVectorD(nk);
299 TMatrixD *y1 = new TMatrixD(3,nk);
300 for (Int_t iClust=0; iClust<nc; iClust++){
301 for(Int_t j=0; j<nSame[iClust]; j++){
302 Int_t iClust1 = cont[iClust][j];
303 for (Int_t iIn=0; iIn<fNeff; iIn++) (*pyx)(iIn,iClust)+=(*mPyx)(iIn,iClust1);
304 (*py)(iClust)+=(*vPy)(iClust1);
305 for (Int_t i=0; i<2; i++) (*y1)(i,iClust)+=(*mY)(i,iClust1);
307 for (Int_t i=0; i<2; i++) (*y1)(i,iClust)/=nSame[iClust];
309 for (Int_t iClust=0; iClust<nk; iClust++) delete [] cont[iClust], delete [] iSame[iClust];
314 for (Int_t iClust=0; iClust<nc; iClust++){
315 for (Int_t iIn=0; iIn<fNeff; iIn++) (*mPyx)(iIn,iClust)=(*pyx)(iIn,iClust);
316 for (Int_t iComp=0; iComp<2; iComp++) (*mY)(iComp,iClust)=(*y1)(iComp,iClust);
317 (*vPy)(iClust)=(*py)(iClust);
320 if (growcl) DoubleClusters(nc,nk,vPy,mY);
329 //-----------------------------------------------------------------------------------
330 void AliDAJetFinder::ReduceClusters(Int_t **iSame,Int_t nc,Int_t &ncout,Int_t **cont,Int_t *nSameOut) const
333 Int_t *nSame = new Int_t[nc];
334 Int_t *iperm = new Int_t[nc];
335 Int_t *go = new Int_t[nc];
336 for (Int_t iCl=0; iCl<nc; iCl++){
338 for (Int_t jCl=0; jCl<nc; jCl++) nSame[iCl]+=iSame[iCl][jCl], cont[iCl][jCl]=0;
342 TMath::Sort(nc,nSame,iperm,true);
344 for (Int_t iCl=0; iCl<nc; iCl++){
348 for (Int_t jCl=0; jCl<nc; jCl++){
349 if (iSame[k][jCl] == 1){
365 //-----------------------------------------------------------------------------------
366 void AliDAJetFinder::EndDetAnn(Int_t &nk,Double_t **xData,Int_t *xx,Double_t etx,TVectorD *vPx,TVectorD *vPy,TMatrixD *mPyx,TMatrixD *mY)
368 //now assign each particle to only one cluster
369 Double_t *clusters=new Double_t[nk];
370 for (Int_t iIn=0; iIn<fNeff; iIn++){
371 for (Int_t iClust=0; iClust<nk; iClust++) clusters[iClust]=(*mPyx)(iIn,iClust);
372 xx[iIn]=TMath::LocMax(nk,clusters);
376 //recalculate codevectors, having all p(y|x)=0 or 1
377 Double_t *xEta = xData[0];
378 Double_t *xPhi = xData[1];
382 for (Int_t iIn=0; iIn<fNin; iIn++){
383 Int_t iClust=xx[iIn];
384 (*mPyx)(iIn,iClust)=1;
385 (*vPy)(iClust)+=(*vPx)(iIn);
386 (*mY)(0,iClust)+=(*vPx)(iIn)*xEta[iIn];
387 (*mY)(3,iClust)+=(*vPx)(iIn)*etx;
390 for (Int_t iClust=0; iClust<nk; iClust++){
391 if ((*vPy)(iClust)>0){
392 Double_t xpos=0,ypos=0,pxy;
393 for (Int_t iIn=0; iIn<fNin; iIn++){
394 pxy=(*vPx)(iIn)*(*mPyx)(iIn,iClust)/(*vPy)(iClust);
395 ypos+=pxy*TMath::Sin(xPhi[iIn]);
396 xpos+=pxy*TMath::Cos(xPhi[iIn]);
397 if (xx[iIn]==iClust) xx[iIn]=k;
399 (*mY)(0,k)=(*mY)(0,iClust)/(*vPy)(iClust);
400 (*mY)(1,k)=(atan2(ypos,xpos)>0) ? atan2(ypos,xpos) : atan2(ypos,xpos)+2*TMath::Pi();
401 (*mY)(3,k)=(*mY)(3,iClust);
408 //-----------------------------------------------------------------------------------
409 void AliDAJetFinder::StoreJets(Int_t nk, Double_t **xData, Int_t *xx, TMatrixD *mY)
411 //evaluate significant clusters properties
412 const Double_t pi=TMath::Pi();
413 AliJetReaderHeader *rHeader=fReader->GetReaderHeader();
414 Float_t dFidEtaMax = rHeader->GetFiducialEtaMax();
415 Float_t dFidEtaMin = rHeader->GetFiducialEtaMin();
416 Float_t dFiducialEta= dFidEtaMax - dFidEtaMin;
417 Double_t *xEta = xData[0];
418 Double_t *xPhi = xData[1];
420 for (Int_t i=0; i<fNeff; i++) if (xEta[i]<dFidEtaMax && xEta[i]>dFidEtaMin) nEff++;
421 Double_t dMeanDist=0.;
423 dMeanDist=TMath::Sqrt(2*dFiducialEta*pi/nEff);
424 Bool_t *isJet = new Bool_t[nk];
425 Double_t *etNoBg= new Double_t[nk];
426 Double_t *dDeltaEta=new Double_t[nk];
427 Double_t *dDeltaPhi=new Double_t[nk];
428 Double_t *surf = new Double_t[nk];
429 Double_t *etDens= new Double_t[nk];
430 for (Int_t iClust=0; iClust<nk; iClust++){
432 Double_t dEtaMin=10.,dEtaMax=-10.,dPhiMin=10.,dPhiMax=0.;
433 for (Int_t iIn=0; iIn<fNeff; iIn++){
434 if (xx[iIn]!=iClust || xEta[iIn]>dFidEtaMax || xEta[iIn]<dFidEtaMin) continue;
435 if (xEta[iIn] < dEtaMin) dEtaMin=xEta[iIn];
436 if (xEta[iIn] > dEtaMax) dEtaMax=xEta[iIn];
437 Double_t dPhi=xPhi[iIn]-(*mY)(1,iClust);
438 if (dPhi > pi ) dPhi-=2*pi;
439 else if (dPhi < (-1)*pi) dPhi+=2*pi;
440 if (dPhi < dPhiMin) dPhiMin=dPhi;
441 else if (dPhi > dPhiMax) dPhiMax=dPhi;
443 dDeltaEta[iClust]=dEtaMax-dEtaMin+dMeanDist;
444 dDeltaPhi[iClust]=dPhiMax-dPhiMin+dMeanDist;
445 surf[iClust]=0.25*pi*dDeltaEta[iClust]*dDeltaPhi[iClust];
446 etDens[iClust]=(*mY)(3,iClust)/surf[iClust];
449 if (((AliDAJetHeader*)fHeader)->GetSelJets()){
450 for (Int_t iClust=0; iClust<nk; iClust++){
451 if (!isJet[iClust] && (*mY)(0,iClust)<dFidEtaMax && (*mY)(0,iClust)>dFidEtaMin){
452 Double_t etDensMed=0.;
453 Double_t etDensSqr=0.;
455 for (Int_t iClust1=0; iClust1<nk; iClust1++){
456 if(iClust1!=iClust && (*mY)(0,iClust)<dFidEtaMax && (*mY)(0,iClust)>dFidEtaMin){
457 etDensMed+=etDens[iClust1];
458 etDensSqr+=TMath::Power(etDens[iClust1],2);
462 etDensMed/=TMath::Max(norm,1);
463 etDensSqr/=TMath::Max(norm,1);
464 Double_t deltaEtDens=TMath::Sqrt(etDensSqr-TMath::Power(etDensMed,2));
465 if ((*mY)(3,iClust) > (etDensMed+deltaEtDens)*surf[iClust]) isJet[iClust]=kTRUE;
466 etNoBg[iClust]=(*mY)(3,iClust)-etDensMed*surf[iClust];
469 for (Int_t iClust=0; iClust<nk; iClust++){
471 Double_t etDensMed=0;
472 Double_t extSurf=2*dFiducialEta*pi;
473 for (Int_t iClust1=0; iClust1<nk; iClust1++){
474 if (!isJet[iClust1]) etDensMed+=(*mY)(3,iClust1);
475 else extSurf-=surf[iClust1];
478 etNoBg[iClust]=(*mY)(3,iClust)-etDensMed*surf[iClust];
479 if (etNoBg[iClust]<((AliDAJetHeader*)fHeader)->GetEtMin()){
480 isJet[iClust]=kFALSE;
486 for (Int_t iClust=0; iClust<nk; iClust++){
488 etNoBg[iClust]=(*mY)(3,iClust);
494 //now add selected jets to the list
495 Int_t *iSort = new Int_t[nk];
496 TMath::Sort(nk,etNoBg,iSort,true);
499 Bool_t fromAod = !strcmp(fReader->ClassName(),"AliJetAODReader");
500 if (fromAod) refs = fReader->GetReferences();
501 for (Int_t iClust=0; iClust<nk; iClust++){ //clusters loop
503 if (isJet[iCl]){ //choose cluster
505 px = (*mY)(3,iCl)*TMath::Cos((*mY)(1,iCl));
506 py = (*mY)(3,iCl)*TMath::Sin((*mY)(1,iCl));
507 pz = (*mY)(3,iCl)/TMath::Tan(2.0 * TMath::ATan(TMath::Exp(-(*mY)(0,iCl))));
508 en = TMath::Sqrt(px * px + py * py + pz * pz);
509 AliAODJet jet(px, py, pz, en);
512 Int_t nEntr = fReader->GetMomentumArray()->GetEntries();
513 for (Int_t iEn=0; iEn<nEntr; iEn++){
514 if (fReader->GetCutFlag(iEn)==0) continue;
515 if (xx[iIn]==iCl) jet.AddTrack(refs->At(iEn));
520 if (fDebug > 0) printf("jet %d, Eta: %f, Phi: %f, Et: %f\n",iCl,jet.Eta(),jet.Phi(),jet.Pt());
523 delete [] dDeltaEta; delete [] dDeltaPhi;
529 //-----------------------------------------------------------------------------------
530 Double_t Dist(TVectorD x,TVectorD y)
533 const Double_t pi=TMath::Pi();
534 Double_t dphi=TMath::Abs(x(1)-y(1));
535 if (dphi > pi) dphi=2*pi-dphi;
536 Double_t dist=pow(x(0)-y(0),2)+pow(dphi,2);