class AliITSTrackleterSPDEff: public AliTracker

 default constructor
 from AliITSMultReconstructor

 from AliITSMultReconstructor

 Copy constructor

 Assignment operator

 Destructor
 from AliITSMultReconstructor
 delete arrays

 - you have to take care of the following, before of using Reconstruct
   1) call LoadClusters(TTree* cl) that finds the position of the clusters (in global coord)
   and  convert the cluster coordinates to theta, phi (seen from the
   interaction vertex).
   2) call SetVertex(vtxPos, vtxErr) which set the position of the vertex
 - Find the extrapolation/interpolation point.
 - Find the chip corresponding to that
 - Check if there is a cluster near that point

 reset counters

 Input: a) layer number in the range [0,1]
        b) vtx[3]: actual vertex
        c) zVtx     \ z of the cluster (-999 for tracklet) computed with respect to vtx
        d) thetaVtx  > theta and phi of the cluster/tracklet computed with respect to vtx
e) phiVtx
 Output: Unique key to locate a chip
 return: kTRUE if succesfull

  Return the average radius of a layer from Geometry

 this method transform Global Cilindrical coordinates into local (i.e. module)
 cartesian coordinates

Compute Cartesian Global Coordinate

 This function finds the detector crossed by the track
 Input: layer in range [0,1]
        phi   in ALICE absolute reference system
         z     "  "       "         "        "

 this method find the intersection in xy between a tracklet (straight line) and
 a circonference (r=R), using polar coordinates.

Input: - vtx[2]: actual vertex w.r.t. ALICE reference system
       - phiVtx: phi angle of the line (tracklet) computed w.r.t. vtx
       - R: radius of the circle
Output: - phi : phi angle of the unique interception in the ALICE Global ref. system

Correct method below: you have the equation of a circle (in polar coordinate) w.r.t. Actual vtx:
r^2-2*r*r0*cos(phi-phi0) + r0^2 = R^2 , where (r0,phi0) is the centre of the circle
In the same system, the equation of a semi-line is: phi=phiVtx;
Hence you get one interception only: P=(r,phiVtx)
Finally you want P in the ABSOLUTE ALICE reference system.

  simple method to reduce all angles (in rad)
  in range [0,2pi[

  This method check if a particle is primary; i.e.
  it comes from the main vertex and it is a "stable" particle, according to
  AliStack::IsPhysicalPrimary() (note that there also Sigma0 are considered as
  a stable particle: it has no effect on this analysis).
  This method can be called only for MC events, where Kinematics is available.
  if fUseOnlyStableParticle is kTRUE (via SetUseOnlyStableParticle) then it
  returns kTRUE if also AliITSTrackleterSPDEff::DecayingTrackChecker() return 0.
  The latter (see below) try to verify if a primary particle is also "detectable".

 This private method can be applied on MC particles (if stack is available),
 provided they have been identified as "primary" from PrimaryTrackChecker() (see above).

 It define "detectable" a primary particle according to the following criteria:

 - if no decay products can be found in the stack (note that this does not
     means it is stable, since a particle is stored in stack if it has at least 1 hit in a
     sensitive detector)
 - if it has at least one decay daughter produced outside or just on the outer pixel layer
 - if the last decay particle is an electron (or a muon) which is not produced in-between
     the two pixel layers (this is likely to be a kink).

 this method allocate memory for the MC related informations
 all the counters are set to 0

 this method deallocate memory for the MC related informations
 all the counters are set to 0

 This method return the Data menmber fPredictionPrimary [1200].
 You can call it only for MC events.
 fPredictionPrimary[key] contains the number of tracklet predictions on the
 given chip key built using  a cluster on the other layer produced (at least)
 from a primary particle.
 Key refers to the chip crossed by the prediction

 This method return the Data menmber fPredictionSecondary [1200].
 You can call it only for MC events.
 fPredictionSecondary[key] contains the number of tracklet predictions on the
 given chip key built using  a cluster on the other layer produced (only)
 from a secondary particle
 Key refers to the chip crossed by the prediction

 This method return the Data menmber fClusterPrimary [1200].
 You can call it only for MC events.
 fClusterPrimary[key] contains the number of tracklet predictions
 built using  a cluster on that layer produced (only)
 from a primary particle
 Key refers to the chip used to build the prediction

 This method return the Data menmber fClusterSecondary [1200].
 You can call it only for MC events.
 fClusterSecondary[key] contains the number of tracklet predictions
 built using  a cluster on that layer produced (only)
 from a secondary particle
 Key refers to the chip used to build the prediction

 This method return the Data menmber fSuccessPP [1200].
 You can call it only for MC events.
 fSuccessPP[key] contains the number of successes (i.e. a tracklet prediction matching
 with a cluster on the other layer) built by using the same primary particle
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fSuccessTT [1200].
 You can call it only for MC events.
 fSuccessTT[key] contains the number of successes (i.e. a tracklet prediction matching
 with a cluster on the other layer) built by using the same  particle (whatever)
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fSuccessS [1200].
 You can call it only for MC events.
 fSuccessS[key] contains the number of successes (i.e. a tracklet prediction matching
 with a cluster on the other layer) built by using a secondary particle
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fSuccessP [1200].
 You can call it only for MC events.
 fSuccessP[key] contains the number of successes (i.e. a tracklet prediction matching
 with a cluster on the other layer) built by using a primary particle
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fFailureS [1200].
 You can call it only for MC events.
 fFailureS[key] contains the number of failures (i.e. a tracklet prediction not matching
 with a cluster on the other layer) built by using a secondary particle
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fFailureP [1200].
 You can call it only for MC events.
 fFailureP[key] contains the number of failures (i.e. a tracklet prediction not matching
 with a cluster on the other layer) built by using a primary particle
 the unique chip key refers to the chip which get updated its efficiency

 This method return the Data menmber fRecons [1200].
 You can call it only for MC events.
 fRecons[key] contains the number of reconstractable tracklets (i.e. a tracklet prediction which
 has an hit in the detector)
 the unique chip key refers to the chip where fall the prediction

 This method return the Data menmber fNonRecons [1200].
 You can call it only for MC events.
 fRecons[key] contains the number of unreconstractable tracklets (i.e. a tracklet prediction which
 has not any hit in the detector)
 the unique chip key refers to the chip where fall the prediction

 Print out some class data values in Ascii Form to output stream
 Inputs:
   ostream *os   Output stream where Ascii data is to be writen
 Outputs:
   none.
 Return:
   none.

 Read in some class data values in Ascii Form to output stream
 Inputs:
   istream *is   Input stream where Ascii data is to be read in from
 Outputs:
   none.
 Return:
   none.

 This Method write into an either asci or root file
 the used cuts and the statistics  of the MC related quantities
 The method SetMC() has to be called before
 Input TString filename: name of file for output (it deletes already existing
 file)
 Output: none


if(!fMC) {CallWarningMC(); return;}

 This Method read from an asci file (do not know why binary does not work)
 the cuts to be used and the statistics  of the MC related quantities
 Input TString filename: name of input file for output
 The method SetMC() has to be called before
 Output: none


if(!fMC) {CallWarningMC(); return;}

 This (private) method save the histograms on the output file
 (only if fHistOn is TRUE).
 Also the histograms from the base class are saved through the
 AliITSMultReconstructor::SaveHists() call

 Saves the histograms into a tree and saves the trees into a file
 Also the histograms from the base class are saved

 This method books addtitional histograms
 w.r.t. those of the base class.
 In particular, the differences of cluster coordinate between the two SPD
 layers are computed in the interpolation phase

 Private method to delete Histograms from memory
 it is called. e.g., by the destructor.

 form AliITSMultReconstructor

 This (private) method can be used only for MC events, where both AliStack and the TrackReference
 are available.
 It is used to asses whether a tracklet prediction is reconstructable or not at the other layer
 Input:
      - Int_t layer (either 0 or 1): layer which you want to check if the tracklete can be
                                     reconstructed at
      - Int_t iC : cluster index used to build the tracklet prediction
                   if layer=0 ==> iC=iC2 ; elseif layer=1 ==> iC=iC1
      - Float_t* vtx: actual event vertex
      - stack: pointer to Stack
      - ref:   pointer to TTRee of TrackReference

 this method apply a rotation by 180 degree around the Z (beam) axis to all
 the RecPoints in a given layer to be used to build tracklets.
 **************** VERY IMPORTANT:: ***************
 It must be called just after LoadClusterArrays, since afterwards the datamember
 fClustersLay1[iC1][0] and fClustersLay1[iC1][1] are redefined using polar coordinate
 instead of Cartesian

 This method is used to find the tracklets.
 It is called from AliReconstruction
 The vertex is supposed to be associated to the Tracker (i.e. to this) already
 The cluster is supposed to be associated to the Tracker already
 In case Monte Carlo is required, the appropriate linking to Stack and TrackRef is attempted

 It is called from AliReconstruction

 This method
 - gets the clusters from the cluster tree
 - convert them into global coordinates
 - store them in the internal arrays
 - count the number of cluster-fired chips

     This method:
  - set Bool_t fLightBackgroundStudyInParallel = b
    a) if you set this kTRUE, then the estimation of the
      SPD efficiency is done as usual for data, but in
      parallel a light (i.e. without control histograms, etc.)
      evaluation of combinatorial background is performed
      with the usual ReflectClusterAroundZAxisForLayer method.
    b) if you set this kFALSE, then you would not have a second
      container for PlaneEfficiency statistics to be used for background
      (fPlaneEffBkg=0). If you want to have a full evaluation of the
      background (with all control histograms and additional data
      members referring to the background) then you have to call the
      method SetReflectClusterAroundZAxisForLayer(kTRUE) esplicitily

 method to study residual background:
 Input b= KTRUE --> reflect the clusters
      ilayer (either 0 or 1) --> which SPD layers should be reflected

 SPD efficiency is done as usual for data, but in
 parallel a light (i.e. without control histograms, etc.)
 evaluation of combinatorial background is performed
 with the usual ReflectClusterAroundZAxisForLayer method.

 for the inner layer

 Only for MC: use only "primary" particles (according to PrimaryTrackChecker) for the tracklet prediction

 Only for MC: use only "secondary" particles (according to PrimaryTrackChecker) for the tracklet prediction

 Only for MC: associate a cluster to the tracklet prediction if  from the same particle

 Only for MC: associate a cluster to the tracklet prediction if  from different particles

  Only for MC: re-define "primary" a particle if it is also "stable" (according to definition in method DecayingTrackChecker)

 only for MC: Getters relative to the above setters

 Getters for the data members related to MC true statisitcs (see below)

 write histograms into a root file on disk

 switch on/off the extra histograms