// @(#)root/eve:$Id: TEveVSDStructs.h 21310 2007-12-10 19:05:45Z matevz $
// Authors: Matevz Tadel & Alja Mrak-Tadel: 2006, 2007

/*************************************************************************
 * Copyright (C) 1995-2007, Rene Brun and Fons Rademakers.               *
 * All rights reserved.                                                  *
 *                                                                       *
 * For the licensing terms see $ROOTSYS/LICENSE.                         *
 * For the list of contributors see $ROOTSYS/README/CREDITS.             *
 *************************************************************************/

#ifndef ROOT_TEveVSDStructs
#define ROOT_TEveVSDStructs

#include "TObject.h"
#include "TMath.h"

#include "TParticle.h"

/******************************************************************************/
// VSD Structures
/******************************************************************************/

// Basic structures for Reve VSD concept. Design criteria:
//
//  * provide basic cross-referencing functionality;
//
//  * small memory/disk footprint (floats / count on compression in
//    split mode);
//
//  * simple usage from tree selections;
//
//  * placement in TClonesArray (composites are TObject derived);
//
//  * minimal member-naming (impossible to make everybody happy).


/******************************************************************************/
// TEveVector
/******************************************************************************/

class TEveVector
{
public:
   Float_t fX, fY, fZ; // Components of the vector.

   TEveVector() : fX(0), fY(0), fZ(0) {}
   TEveVector(Float_t x, Float_t y, Float_t z) : fX(x), fY(y), fZ(z) {}
   virtual ~TEveVector() {}

   TEveVector operator + (const TEveVector &);
   TEveVector operator - (const TEveVector &);
   TEveVector operator * (Float_t a);

   Float_t& operator [] (Int_t indx);
   Float_t  operator [] (Int_t indx) const;

   const Float_t* Arr() const { return &fX; }
         Float_t* Arr()       { return &fX; }

   void Set(Float_t*  v) { fX = v[0]; fY = v[1]; fZ = v[2]; }
   void Set(Double_t* v) { fX = v[0]; fY = v[1]; fZ = v[2]; }
   void Set(Float_t  x, Float_t  y, Float_t  z) { fX = x; fY = y; fZ = z; }
   void Set(Double_t x, Double_t y, Double_t z) { fX = x; fY = y; fZ = z; }
   void Set(const TVector3& v)   { fX = v.x(); fY = v.y(); fZ = v.z(); }
   void Set(const TEveVector& v) { fX = v.fX;  fY = v.fY;  fZ = v.fZ;  }

   Float_t Phi()      const;
   Float_t Theta()    const;
   Float_t CosTheta() const;
   Float_t Eta()      const;

   Float_t Mag()  const { return TMath::Sqrt(fX*fX + fY*fY + fZ*fZ);}
   Float_t Mag2() const { return fX*fX + fY*fY + fZ*fZ;}

   Float_t Perp()  const { return TMath::Sqrt(fX*fX + fY*fY);}
   Float_t Perp2() const { return fX*fX + fY*fY;}
   Float_t R()     const { return Perp(); }

   Float_t Distance(const TEveVector& v) const;
   Float_t SquareDistance(const TEveVector& v) const;
   Float_t Dot(const TEveVector&a) const;

   TEveVector& Mult(const TEveVector&a, Float_t af)
   { fX = a.fX*af; fY = a.fY*af; fZ = a.fZ*af; return *this; }


   ClassDef(TEveVector, 1); // Float three-vector; a inimal Float_t copy of TVector3 used to represent points and momenta (also used in VSD).
};

//______________________________________________________________________________
inline Float_t TEveVector::Phi() const
{ return fX == 0.0 && fY == 0.0 ? 0.0 : TMath::ATan2(fY, fX); }

inline Float_t TEveVector::Theta() const
{ return fX == 0.0 && fY == 0.0 && fZ == 0.0 ? 0.0 : TMath::ATan2(Perp(), fZ); }

inline Float_t TEveVector::CosTheta() const
{ Float_t ptot = Mag(); return ptot == 0.0 ? 1.0 : fZ/ptot; }

inline Float_t TEveVector::Distance( const TEveVector& b) const
{
   return TMath::Sqrt((fX - b.fX)*(fX - b.fX) +
                      (fY - b.fY)*(fY - b.fY) +
                      (fZ - b.fZ)*(fZ - b.fZ));
}
inline Float_t TEveVector::SquareDistance(const TEveVector& b) const
{
   return ((fX - b.fX) * (fX - b.fX) +
           (fY - b.fY) * (fY - b.fY) +
           (fZ - b.fZ) * (fZ - b.fZ));
}

//______________________________________________________________________________
inline Float_t TEveVector::Dot(const TEveVector& a) const
{
   return a.fX*fX + a.fY*fY + a.fZ*fZ;
}

inline Float_t& TEveVector::operator [] (Int_t idx)
{ return (&fX)[idx]; }

inline Float_t TEveVector::operator [] (Int_t idx) const
{ return (&fX)[idx]; }


/******************************************************************************/
// TEvePathMark
/******************************************************************************/

class TEvePathMark
{
public:
   enum EType_e   { kReference, kDaughter, kDecay };

   TEveVector  fV;    // Vertex.
   TEveVector  fP;    // Momentum.
   Float_t     fTime; // Time.
   EType_e     fType; // Mark-type.

   TEvePathMark(EType_e t=kReference) : fV(), fP(), fTime(0), fType(t) {}
   virtual ~TEvePathMark() {}

   const char* TypeName();

   ClassDef(TEvePathMark, 1); // Special-point on track: position/momentum reference, daughter creation or decay (also used in VSD).
};

/******************************************************************************/
// TEveMCTrack
/******************************************************************************/

class TEveMCTrack : public TParticle // ?? Copy stuff over ??
{
public:
   Int_t       fLabel;      // Label of the track
   Int_t       fIndex;      // Index of the track (in some source array)
   Int_t       fEvaLabel;   // Label of primary particle

   Bool_t      fDecayed;    // True if decayed during tracking.
   // ?? Perhaps end-of-tracking point/momentum would be better.
   Float_t     fTDecay;     // Decay time
   TEveVector  fVDecay;     // Decay vertex
   TEveVector  fPDecay;     // Decay momentum

   TEveMCTrack() : fLabel(-1), fIndex(-1), fEvaLabel(-1),
                   fDecayed(kFALSE), fTDecay(0), fVDecay(), fPDecay() {}
   virtual ~TEveMCTrack() {}

   TEveMCTrack& operator=(const TParticle& p)
   { *((TParticle*)this) = p; return *this; }

   void ResetPdgCode() { fPdgCode = 0; }

   ClassDef(TEveMCTrack, 1); // Monte Carlo track (also used in VSD).
};


/******************************************************************************/
// TEveHit
/******************************************************************************/

// Representation of a hit.

// Members det_id (and fSubdetId) serve for cross-referencing into
// geometry. Hits should be stored in fDetId (+some label ordering) in
// order to maximize branch compression.


class TEveHit : public TObject
{
public:
   UShort_t     fDetId;    // Custom detector id.
   UShort_t     fSubdetId; // Custom sub-detector id.
   Int_t        fLabel;    // Label of particle that produced the hit.
   Int_t        fEvaLabel; // Label of primary particle, ancestor of label.
   TEveVector   fV;        // Hit position.

   // Float_t charge; probably specific.

   TEveHit() : fDetId(0), fSubdetId(0), fLabel(0), fEvaLabel(0), fV() {}
   virtual ~TEveHit() {}

   ClassDef(TEveHit, 1); // Monte Carlo hit (also used in VSD).
};


/******************************************************************************/
// TEveCluster
/******************************************************************************/

// Base class for reconstructed clusters

// ?? Should TEveHit and cluster have common base? No.

class TEveCluster : public TObject
{
public:
   UShort_t     fDetId;     // Custom detector id.
   UShort_t     fSubdetId;  // Custom sub-detector id.
   Int_t        fLabel[3];  // Labels of particles that contributed hits.

   // ?? Should include reconstructed track(s) using it? Rather not, separate.

   TEveVector      fV;      // Vertex.
   // TEveVector   fW;      // Cluster widths.
   // Coord system? Errors and/or widths Wz, Wy?

   TEveCluster() : fDetId(0), fSubdetId(0), fV() { fLabel[0] = fLabel[1] = fLabel[2] = 0; }
   virtual ~TEveCluster() {}

   ClassDef(TEveCluster, 1); // Reconstructed cluster (also used in VSD).
};


/******************************************************************************/
// TEveRecTrack
/******************************************************************************/

class TEveRecTrack : public TObject
{
public:
   Int_t       fLabel;       // Label of the track.
   Int_t       fIndex;       // Index of the track (in some source array).
   Int_t       fStatus;      // Status as exported from reconstruction.
   Int_t       fSign;        // Charge of the track.
   TEveVector  fV;           // Start vertex from reconstruction.
   TEveVector  fP;           // Reconstructed momentum at start vertex.
   Float_t     fBeta;        // Relativistic beta factor.

   // PID data missing

   TEveRecTrack() : fLabel(-1), fIndex(-1), fStatus(0), fSign(0), fV(), fP(), fBeta(0) {}
   virtual ~TEveRecTrack() {}

   Float_t Pt() { return fP.Perp(); }

   ClassDef(TEveRecTrack, 1); // Reconstructed track (also used in VSD).
};


/******************************************************************************/
// TEveRecKink
/******************************************************************************/

class TEveRecKink : public TEveRecTrack
{
public:
   Int_t       fLabelSec;  // Label of the secondary track
   TEveVector  fVEnd;      // End vertex: last point on the primary track
   TEveVector  fVKink;     // Kink vertex: reconstructed position of the kink
   TEveVector  fPSec;      // Momentum of secondary track

   TEveRecKink() : TEveRecTrack(), fLabelSec(0), fVEnd(), fVKink(), fPSec() {}
   virtual ~TEveRecKink() {}

   ClassDef(TEveRecKink, 1); // Reconstructed kink (also used in VSD).
};


/******************************************************************************/
// TEveRecV0
/******************************************************************************/

class TEveRecV0 : public TObject
{
public:
   Int_t      fStatus;

   TEveVector fVNeg;       // Vertex of negative track.
   TEveVector fPNeg;       // Momentum of negative track.
   TEveVector fVPos;       // Vertex of positive track.
   TEveVector fPPos;       // Momentum of positive track.

   TEveVector fVCa;        // Point of closest approach.
   TEveVector fV0Birth;    // Reconstucted birth point of neutral particle.

   // ? Data from simulation.
   Int_t      fLabel;      // Neutral mother label read from kinematics.
   Int_t      fPdg;        // PDG code of mother.
   Int_t      fDLabel[2];  // Daughter labels.

   TEveRecV0() : fStatus(), fVNeg(), fPNeg(), fVPos(), fPPos(),
                 fVCa(), fV0Birth(), fLabel(0), fPdg(0)
   { fDLabel[0] = fDLabel[1] = 0; }
   virtual ~TEveRecV0() {}

   ClassDef(TEveRecV0, 1); // Reconstructed V0 (also used in VSD).
};

/******************************************************************************/
/******************************************************************************/

// Missing primary vertex.

// Missing TEveMCRecCrossRef, RecInfo.

class TEveMCRecCrossRef : public TObject
{
public:
   Bool_t       fIsRec;   // Is reconstructed.
   Bool_t       fHasV0;
   Bool_t       fHasKink;
   Int_t        fLabel;
   Int_t        fNHits;
   Int_t        fNClus;

   TEveMCRecCrossRef() : fIsRec(false), fHasV0(false), fHasKink(false),
                         fLabel(0), fNHits(0), fNClus(0) {}
   virtual ~TEveMCRecCrossRef() {}

   ClassDef(TEveMCRecCrossRef, 1); // Cross-reference of sim/rec data per particle (also used in VSD).
};

/******************************************************************************/
/******************************************************************************/

// This whole construction is somewhat doubtable. It requires
// shameless copying of experiment data. What is good about this
// scheme:
//
// 1) Filters can be applied at copy time so that only part of the
// data is copied over.
//
// 2) Once the data is extracted it can be used without experiment
// software. Thus, external service can provide this data and local
// client can be really thin.
//
// 3) Some pretty advanced visualization schemes/selections can be
// implemented in a general framework by providing data extractors
// only. This is also good for PR or VIP displays.
//
// 4) These classes can be extended by particular implementations. The
// container classes will use TClonesArray with user-specified element
// class.

// The common behaviour could be implemented entirely without usage of
// a common base classes, by just specifying names of members that
// retrieve specific data. This is fine as long as one only uses tree
// selections but becomes painful for extraction of data into local
// structures (could a) use interpreter but this is an overkill and
// would cause serious trouble for multi-threaded environment; b) use
// member offsets and data-types from the dictionary).

#endif

Last update: Thu Jan 17 08:48:44 2008

This page has been automatically generated. If you have any comments or suggestions about the page layout send a mail to ROOT support, or contact the developers with any questions or problems regarding ROOT.