class AliMultiplicityCorrection: public TNamed

 sets the quality region definition to TPC or SPD

 default constructor

 opens the given file, reads the multiplicity from the given folder and returns the object

 named constructor

 rebins the y axis of a two-dimensional histogram giving variable size binning (missing in ROOT v5/25/02)

 rebins the y axis of a three-dimensional histogram giving variable size binning (missing in ROOT v5/25/02)
 this function is a mess - and it should have been Fons who should have gone through the pain of writing it! (JF)

 Rebins the (and only the) generated multiplicity axis

 Destructor

 Merge a list of AliMultiplicityCorrection objects with this (needed for
 PROOF).
 Returns the number of merged objects (including this).

 loads the histograms from a file
 if dir is empty a directory with the name of this object is taken (like in SaveHistogram)

 saves the histograms

 Fills an event from MC

 Fills an event from ESD

 fills raw distribution of triggered events

 fills raw distribution of triggered events

 Fills an event into the correlation map with the information from MC and ESD

 fills fCurrentESD, fCurrentCorrelation
 resets fMultiplicityESDCorrected

 calculates efficiency for given event type

 calculates efficiency for given event type

 correct spectrum using minuit chi2 method

 for description of parameters, see AliUnfolding::Unfold

 fills the 0 bin

 sets trigger and vertex efficiencies to 1 for large multiplicities where no event was simulated

 returns the fraction of events that have 0 generated particles in the given range among all events without vertex OR 0 tracklets

 correct spectrum using minuit chi2 method with a NBD function

 for description of parameters, see AliUnfolding::Unfold

 draws the histograms of this class

 draw comparison plots

   This computes an in-place complex-to-complex FFT
   x and y are the real and imaginary arrays of 2^m points.
   dir =  1 gives forward transform
   dir = -1 gives reverse transform

     Formula: forward
                  N-1
                  ---
              1   \          - j k 2 pi n / N
      X(n) = ---   >   x(k) e                    = forward transform
              N   /                                n=0..N-1
                  ---
                  k=0

      Formula: reverse
                  N-1
                  ---
                  \          j k 2 pi n / N
      X(n) =       >   x(k) e                    = forward transform
                  /                                n=0..N-1
                  ---
                  k=0

 Returns the chi2 between the MC and the unfolded ESD as well as between the ESD and the folded unfolded ESD
 These values are computed during DrawComparison, thus this function picks up the
 last calculation

 returns the corresponding MC spectrum

 returns the corresponding MC spectrum

 calculate standard deviation of (results[0] - results[k]) k=1...max-1
 per bin one gets: sigma(r[0] - r[n]) / r[0]

 evaluates the uncertainty that arises from the non-infinite statistics in the response matrix
 the function unfolds the spectrum using the default response matrix and several modified ones
 the modified ones are created by randomizing each cell using poisson statistics with the mean = bin value
 these unfolded results are compared to the first result gained with the default response OR to the histogram given
 in <compareTo> (optional)

 returns the error assigned to the measurement

 correct spectrum using bayesian method

 determineError:
   0 = no errors
   1 = from randomizing
   2 = with UnfoldGetBias

 helper function for the covariance matrix of the bayesian method

 correct spectrum using bayesian method

 correct spectrum using a simple Gaussian approach, that is model-dependent

 convolutes the corrected histogram i with the response matrix

 creates the residual histogram from the corrected histogram i corresponding to an eventType event sample using the corresponding correlation matrix
 residual is : M - UT / eM
 residualSum contains the squared sum of the residuals

 runs the distribution given in inputMC through the response matrix identified by
 correlationMap and produces a measured distribution
 although it is a TH2F the vertex axis is not used at the moment and all entries are filled in mid-vertex

 uses the given function to fill the input MC histogram and generates from that
 the measured histogram by applying the response matrix
 this can be used to evaluate if the methods work indepedently of the input
 distribution
 WARNING does not respect the vertex distribution, just fills central vertex bin