class AliAnalysisTaskJetV2: public AliAnalysisTaskEmcalJet

 destructor

 Init the analysis

 determine the run number to see if the track and jet cuts should be refreshed for semi-good TPC runs

 initialize the anaysis

 book a TH1F and connect it to the output container

 book a TH2F and connect it to the output container

 book a TH2F and connect it to the output container

 create output objects. also initializes some default values in case they aren't
 loaded via the AddTask macro

 called for each accepted event (call made from user exec of parent class)

 for stand alone, avoid framework event setup

 numerically integrate with finite resolution
 idea is the following:
 1) choose a vector phi
 2) see if it is in a region of overlap between detector and in/out of plane spectrum
 3) bookkeep percentages over overlap

 'numerical integration' of geometric overlap

 works as follows: for a given vector phi determines whether
 or not this vector points towards an overlap region of
 detector geometry and plane (in or out)

 returns
 1) if overlap with in plane
 2) if overlap with out of plane
 0) if no overlap at all

 return chi for given resolution to combine event plane estimates from two subevents
 see Phys. Rev. C no. CS6346 (http://arxiv.org/abs/nucl-ex/9805001)

 get the vzero event plane (a and c separately)

 return the combined vzero event plane

 grab the TPC event plane

 fill the profiles for the resolution parameters

 return the calibrated 2nd and 3rd order q-vectors for vzeroa and vzeroc
 function takes arrays as arguments, which correspond to vzero info in the following way

 Qa2[0] = Qx2 for vzero A         Qa2[1] = Qy2 for vzero A (etc)

 calculate calibrated q-vector of the combined vzeroa, vzeroc system
 this is somewhat ugly as CalculateQvectorCombinedVZERO is called more than once per event
 but for now it will have to do ...

 get a random cone

 get the second order q-cumulant, a -999 return will be caught in the qa routine of CorrectRho

 get the fourth order q-cumulant, a -999 return will be caught in the qa routine of CorrectRho

 get the weighted n-th order q-vector, pass real and imaginary part as reference

 get  unweighted differential flow vectors

 get the weighted ij-th order autocorrelation correction

 get multiplicity for unweighted q-cumulants. function QCnQnk should be called first

 get multiplicity weights for the weighted two particle cumulant

 get multiplicity weights for the weighted four particle cumulant

 decides how to deal with the situation where c2 or c3 is negative
 returns kTRUE depending on whether or not a modulated rho is used for the jet background

 get rho' -> rho(phi)
 two routines are available, both can be used with or without pt weights
  [1] get vn from q-cumulants or as an integrated value from a user supplied histogram
      in case of cumulants, both cumulants and vn values are stored. in both cases, v2 and v3
      are expected. a check is performed to see if rho has no negative local minimum
      for full description, see Phys. Rev. C 83, 044913
      since the cn distribution has negative values, vn = sqrt(cn) can be imaginary sometimes
      in this case one can either roll back to the 'original' rixed rho, do a fit for vn or take use
      vn = - sqrt(|cn|)
  [2] fitting a fourier expansion to the de/dphi distribution
      the fit can be done with either v2, v3 or a combination.
      in all cases, a cut can be made on the p-value of the chi-squared value of the fit
      and a check can be performed to see if rho has no negative local minimum

 event cuts

 fill histograms

 fill qa histograms for pico tracks

 fill qa histograms for events

 fill track histograms

 fill cluster histograms

 fill event plane histograms, only called in qa mode

 fill rho histograms

 fill delta pt histograms

 fill jet histograms

 fill qa histograms for pico tracks

 fill qa histograms for events

 fill the trigger efficiency histograms

 fill the analysis summary histrogram, saves all relevant analysis settigns

 terminate

 set modulation fit

 set control fit

 INTERFACE METHOD FOR OUTPUTFILE
 get the detector resolution, user has ownership of the returned histogram

 INTERFACE METHOD FOR OUTPUT FILE
 correct the supplied differential vn histogram v for detector resolution

 INTERFACE METHOD FOR OUTPUT FILE
 correct the supplied intetrated vn histogram v for detector resolution
 integrated vn must have the same centrality binning as the resolotion correction

 get differential QC

 necessary for calibration of 10h vzero event plane. code copied from flow package
 (duplicate, but i didn't want to introduce an ulgy dependency )
 this function is only called when the runnumber changes

 return cache index number corresponding to the event centrality

 return pointer to the highest pt jet (before background subtraction) within acceptance
 only rudimentary cuts are applied on this level, hence the implementation outside of
 the framework

 get event weights distribution from event plane distribution

 test function to print binary representation of given trigger mask
 trigger mask is represented by 32 bits (hardcoded as it is an UInt_t )

 constructors, destructor

 setters - analysis setup

 setters specific to the vzero calibration for 10h data

 exclude vzero rings: 0 through 7 can be excluded by calling this setter multiple times
 0 corresponds to segment ID 0 through 7, etc

 getters

 local cuts

 numerical evaluations

 filling histograms

 interface methods for the output file