class AliTRDCalibraFillHisto: public TObject

 Singleton implementation

 Singleton implementation
 Deletes the instance of this class

 Default constructor

 Copy constructor

 Delete instance

 Delete DebugStreamer

 Delete the histos

 Init the histograms and stuff to be filled

 Init the Gain Cal Det

 Init the Gain Cal Pad

 Use AliTRDtrackV1 for the calibration

 Drift velocity calibration:
 Fit the clusters with a straight line
 From the slope find the drift velocity

 PRF width calibration
 Assume a Gaussian shape: determinate the position of the three pad clusters
 Fit with a straight line
 Take the fitted values for all the clusters (3 or 2 pad clusters)
 Fill the PRF as function of angle of the track

 See if we are not near a masked pad

 See if we are not near a masked pad

 Look in the choosen database if the pad is On.
 If no the track will be "not good"

 Calculate the calibration group number for i

 Calculate the total number of calibration groups

 Set the mode of calibration group in the z direction for the parameter i

 Set the mode of calibration group in the rphi direction for the parameter i

 Set the mode of calibration group all together

 Set the mode of calibration group per supermodule

 For the detector calcul the first Xbins and set the number of row
 and col pads per calibration groups, the number of calibration
 groups in the detector.

 Should be between 0 and 6

 Store the infos in fAmpTotal, fPHPlace and fPHValue
 Correct from the gain correction before
 cls is shared cluster if any
 Return the charge

 Reset values per tracklet

 For the offline tracking or mcm tracklets
 This function will be called in the functions UpdateHistogram...
 to fill the info of a track for the relativ gain calibration

 For the offline tracking or mcm tracklets
 This function will be called in the functions UpdateHistogram...
 to fill the info of a track for the drift velocity  calibration

 Event Processing loop - AliTRDrawStream

 0 timebin problem
 1 no input
 2 input
 Same algorithm as TestBeam but different reader

 For the DAQ
 Fill a simple average pulse height

  Write infos to file

 Check the number of stats in h, 0 is TH2I 1 is TProfile2D
 debug mode with 2 for TH2I and 3 for TProfile2D
 It gives a pointer to a Double_t[7] with the info following...
 [0] : number of calibration groups with entries
 [1] : minimal number of entries found
 [2] : calibration group number of the min
 [3] : maximal number of entries found
 [4] : calibration group number of the max
 [5] : mean number of entries found
 [6] : mean relative error

 Return a Int_t[4] with:
 0 Mean number of entries
 1 median of number of entries
 2 rms of number of entries
 3 number of group with entries

 Return a Int_t[4] with:
 0 Mean number of entries
 1 median of number of entries
 2 rms of number of entries
 3 number of group with entries

 Create the 2D histos: here we have 2*fNgroupprf bins in tnp of 0.2 amplitude each
 If fNgroupprf is zero then no binning in tnp

 Create the 2D histos

 Create the 2D histos

 Set the factor that will divide the deposited charge
 to fit in the histo range [0,fRangeHistoCharge]

 FillCH2d: Marian style

 Reconstruct the layer number from the detector number

 Reconstruct the stack number from the detector number

 Reconstruct the sector number from the detector number

 return pointer to fPH2d TProfile2D
 create a new TProfile2D if it doesn't exist allready

 return pointer to fCH2d TH2I
 create a new TH2I if it doesn't exist allready

 return pointer to TLinearFitter Calibration
 if force is true create a new TLinearFitter if it doesn't exist allready

 Analyse array of linear fitter because can not be written
 Store two arrays: one with the param the other one with the error param + number of entries

 Set of Get the variables

 Choice to fill or not the 2D

 Debug

 Vector method