MUEX

The muex module of mue implements the most recent version of the in-ice track and cascade energy reconstruction of the project mue. A likelihood function consists of both statistical (Poisson) and systematic (describing errors in the photon flux description) terms.

Photon parametrization relies on ice files icemodel.dat and icemodel.par that describe the ice scattering and absorption coefficients. These should be contained within the directory specified with parameter icedir. Optionally if files tilt.dat and tilt.par exist they will be used to describe the tilt of the ice layers. These files are exactly the same as in the ice description used by direct photon propagators (e.g., ppc). For best results most recent ice description should be used, but different ice models will obviously lead to a varition in the result.

The systematic part of the likelihood function describing the errors of the photon parametrization is a wider log-normal gaussian (a 1-degree student’s t-distribution) when used for cascade reconstruction of in a compatibility mode for track reconstruction. If the compatibility mode is not used the newer skewed description is used in track reconstruction that allows for large over-fluctuations (that happen in a form of, e.g., bremsstrahlung losses of muons). This improves resolution in the muon energy reconstruction to the levels comparable with TE.

In detailed mode a particle vector is produced that containes reconstructed losses at equal distance intervals along the muon track, similar to what is done by millipede. Since only time-integrated analytical description of the photon flux is used, the precision is obviously lower than that achieved by millipede, with an error level of less than about 10% in both the described charge deposition and total energy deposition.

If the input track name is empty the module will calculate its own track, and as such can be used as a simple-and-quick stand-alone track/cascade reconstuction tool. In this mode use parameter usempe to switch between the SPE (using the full time-series) and MPE (using only the first photon and total charge in each DOM) descriptions.

If a parameter repeat is set then the input recopulse series will be re-sampled according to a multinomial distribution (preserving total charge in the event) and uncertainties in some parameters (energy, angular resolution if rectrk is empty) will be calculated by a bootrapping resampling method. Additionally, the precision of the reconstruction will usually go up with more resamplings as the result of each iteration is checked against the full recopulse series to see if it yields an improved solution.

Finally, this module is not a toy. Please use responsibly!

Usage

Usage is straightforward. Available parameters are listed below (from icetray-inspect mue):

muex
  badoms
    Description : list of clipped/saturated OMs
    Default     : ''

  compat
    Description : compatibility with older pre-wreco trunk
    Default     : False

  detail
    Description : detailed energy losses
    Default     : False

  energy
    Description : estimate energy
    Default     : False

  icedir
    Description : icemodel directory
    Default     : '$I3_BUILD/mue/resources/ice/mie'

  lcspan
    Description : lc span in input data
    Default     : 0

  pulses
    Description : input pulse series
    Default     : 'InPulses'

  rectrk
    Description : input track name
    Default     : ''

  rectyp
    Description : track(True) or cascade(False) reconstruction
    Default     : True

  repeat
    Description : number of iterations for uncertainty estimatation
    Default     : 0

  result
    Description : name of the result particle
    Default     : 'MuEx'

  usempe
    Description : mpe(True) or spe(False) reconstruction
    Default     : False

Output

This module can write several output objects to the frame.

  • result: An I3Particle. The position, time and direction of this particle will either be copied from ‘rectrk, if it exists, or otherwise calculated by this module. If the ‘energy’ option is set, this module will also compute an energy estimate which will be stored as the particle’s energy.

  • result + “_r”: Written only when ‘detail’ is set to True. A number indicating the ratio of charge predicted by the energy loss unfolding to the actually observed total charge.

  • result + “_list”: Written only when ‘detail’ is set to True. A list of particles representing the energy losses which were unfolded.

  • result + “_rllt”: Only written when ‘rectrk’ is not set, so this module does its own reconstruction. A number representing the likelihood value obtained from the directional/postional fit.

  • result + “_rlle”: Only written when ‘rectrk’ is not set, so this module does its own reconstruction, and ‘energy’ is set to True. A number representing the likelihood value obtained from the energy fit.

  • result + “_Sigma”: Only written when ‘rectrk’ is not set, so this module does its own reconstruction, and ‘repeat’ is larger than zero. A number representing the median angular difference, in radians, between the bootstrapped repetitions of the directional reconstruction and the average of those reconstructions. This can be interpreted as a measure of the uncertainty of the directional fit.

  • result + “_EnUnc”: Only written when ‘rectrk’ is not set, so this module does its own reconstruction, ‘energy’ is set to True, and ‘repeat’ is larger than zero. A number representing the median absolute value of the difference between the natural logarithm of the energies obtained from the bootstrapped repetitions of the energy reconstruction and the average natural logarithm of those energies. This can be interpreted as a measure of the uncertainty of the energy estimate.