class AliAnalysisTaskFilteredTree: public AliAnalysisTaskSE

 Constructor

 Destructor

 Create histograms
 Called once

 Called for each event

 Find cosmic pairs (triggered or random)

 Select real events with high-pT tracks

 Process laser events -> dump tracks and clusters  to the special tree

 Select real events with high-pT tracks
 Calculate and stor in the output tree:
  TPC constrained tracks
  InnerParams constrained tracks
  TPC-ITS tracks
  ITSout-InnerParams tracks
  chi2 distance between TPC constrained and TPC-ITS tracks
  chi2 distance between TPC refitted constrained and TPC-ITS tracks
  chi2 distance between ITSout and InnerParams tracks
  MC information:
   track references at ITSin, TPCin; InnerParam at first TPC track reference,
   particle ID, mother ID, production mechanism ...

 get selection cuts

 Fill tree for efficiency studies MC only

 check if particle is Z > 1

 Select real events with V0 (K0s and Lambda and Gamma) high-pT candidates

 Select real events with large TPC dEdx signal

 Create KF particle in case the V0 fullfill selection criteria

 Selection criteria
  0. algorithm cut
  1. track cut
  3. chi2 cut
  4. rough mass cut
  5. Normalized pointing angle cut

 Downscale randomly low pt V0

return kFALSE;

 Constrain TPC inner params constrained

 Constrain TPC inner params constrained

 Called one at the end
 locally on working node

 Called one at the end

 calculate mc event true track multiplicity

 Fill pT relative resolution histograms for
 TPC only, TPC only constrained to vertex and TPC+ITS tracking

 Process ITS standalone tracks find match with closest TPC(or combined tracks) tracks
 marian.ivanov@cern.ch
 0.) Init variables
 1.) GetTrack parameters at TPC inner wall
 2.) Match closest TPC  track  (STANDALONE/global) - chi2 match criteria

 Logic to be used in reco:
 1.) Find matching ITSalone->TPCalone
 2.) if (!TPCalone.FindClose(TPCother))  TPCalone.Addopt(ITSalone)
 3.) ff ((ITSalone.FindClose(Global)==0) CreateGlobaltrack

    Use TPC standalone, ITS standalone and combined tracks to categorize/resp. recover track information.

    Track categories:
       -TPC+ITS
       -TPC only
       -ITS only
    Topology categories:
       -Nice isolated tracks with full information
       -Overlapped tracks - Refit and sign them
       -Multiple found (check overlap factor) - Merge and sign
       -Charge particle (kink) decays - Change of direction - Sign them)
    Info:
       -Array  of indexes of closest tracks in each individual category
       -Chi2  distance to the closest tracks at reference radius of TPCin
       -Overlap factors  - fraction of shared clusters or missing  region
       -Chi2 distance at DCA
    Information matrix:
       -matrix closest tracks from each categories
       -characterization - chi2, index,chi2,  overlap ratio

    Decissions:
       0.) Kink decay or catastophic multiple scaterring
           (combining all track categories)
              - small chi2 at the DCA
              - significantly large deflection angle
              - Combinatorial algorithm - to decrease CPU time restriction of investigation to tracks with small DCA at  refernce radius

       1.) if (TPConly && !(TPC+ITS) && ITSonly match ) {
             if (!kink) TPCOnly.addoptITS
             if (kink) TPConly sign
           }

       2.) Overlap tracks - Refit with doUnfold