Xmax data from the Pierre Auger Observatory as published in

  A. Aab et al. [Pierre Auger Collaboration], XXX Y (2014), ZZZ


files:

  example.C
  =========

  ROOT macro to examplify the reading of the various files.


  xmaxMoments.txt
  ===============
  mean and standard deviation of the Xmax distribution (fully
  corrected for detector effects).

  Format:
    energyBin lgMinEnergy lgMaxEnergy meanLgEnergy numberOfEvents
    meanXmax meanXmaxSigmaStat meanXmaxSigmaSysUp meanXmaxSigmaSysLow
    sigmaXmax sigmaXmaxSigmaStat sigmaXmaxSigmaSysUp
    sigmaXmaxSigmaSysLow

  xmaxHistograms.txt
  ==================
  This file contains the number of entries per Xmax interval for
  different energies.

  Format:
    energyBin lgMinEnergy lgMaxEnergy numberOfEvents numberOfBins xmaxMin xmaxMax
  followed by numberOfBins numbers denoting the entries in the respective bin.


  xmaxSystematics.txt
  ===================
  Table of Xmax-scale uncertainty.

  Format:
    energyBin lgMinEnergy lgMaxEnergy sigmaUp sigmaLow

  sigmaUp and sigmaLow are the asymetric uncertainties in units of g/cm^2.


  correlationSysPlus/Minus.txt
  ============================
  Correlation matrix of the Xmax-scale uncertainty stored as a 18x18 table.


  resolution.txt
  ==============
  Table of parameters of Xmax resolution as a function of energy.
  The resolution is a sum of Gaussians G(mean, sigma)
     f(Xmax_true - Xmax_rec) = k * G(0, sigma1) + (1 - k) * G(0, sigma2)
  with variance
     V = k * sigma1^2 + (1 - k) * sigma2^2

  Format:
    energyBin lgMinEnergy lgMaxEnergy sigma1 sigma1Err sigma2 sigma2Err k

  sigma1 and sigma2 and their syst. errors are in units of g/cm^2.
  Note that sigma1Err and sigma2Err are correlated, i.e. the upper
  and lower 1-sigma bound on the resolution is given by
    k * G(0, sigma1 + sigma1Err) + (1 - k) * G(0, sigma2 + sigma2Err)
  and
    k * G(0, sigma1 - sigma1Err) + (1 - k) * G(0, sigma2 - sigma2Err)
  respectively.


  acceptance.txt
  ==============
  Table or parameters of the relative Xmax acceptance,

              /  exp((Xmax-x1)/l1,   Xmax <= x1
    relAcc = <   1,                  x1 < Xmax <= x2
              \  exp(-(Xmax-x2)/l2), Xmax > x2

  Format:
    energyBin lgMinEnergy lgMaxEnergy x1 ex1 x2 ex2 l1 el1 l2 el2
  where ex1, ex2, el1 and el2 are the uncertainties of the parameters.

  All acceptance parameters are in units of g/cm^2.
  Note that the acceptance is parameterized as a function of *true* Xmax,
  i.e. before the detector resolution.
  Extreme cases of the acceptance can be obtained by using
     (x1-ex1, x2+ex2, l1+el1, l2+el2)
  and
     (x1+ex1, x2-ex2, l1-el1, l2-el2)
  as parameters.
