cra-ahlers-llh

• Iterative log-likelyhood reconstruction for CRA
• C++ implementation of maximum-likelihood technique for reconstructing cosmic-ray anisotropy maps
• Cite: M. Ahlers et al 2016 ApJ 823 10

scripts

Collection of scripts for driving production of extraction, reconstruction and analysis of cosmic-ray data.

Installation

Prerequisites:

iter-lhreco dependencies:

• CFITSIO: depends on cfitsio: https://heasarc.gsfc.nasa.gov/fitsio/fitsio.html
• GNU Scientific Library: depends on gsl: https://www.gnu.org/software/gsl/doc/html/index.html
• HEALpix: depends on healpix: http://healpix.sourceforge.net/
• BOOST C++ libraries: https://www.boost.org/

You will also need:

• CMake >= 3.1
• A C++11-compliant compiler (e.g. gcc >= 4.8.1 or clang >= 3.3)

Installation:

The C++ projects are built using CMake. Each project has a CMakeList.txt that will detect dependencies and generate a MakeFile. To do this cd into the build directory and excecute the commands:

cmake ../src; make

Running:

You need to provide a JSON configuration file that lists the parameters for each observatory. There is an examples directory that contains configurations for IceCube, HAWC and IceCube+HAWC. The config file can contain an arbitrary number of observatories.

The output will be in the form of a HEALPix map of NSide = 64 but you can specify a different NSide. If you don't specify the number of iterations, it will default to 20.

Other parameters such as --seed, --iso, and --fluctuate are only used for systematic studies.

  ./bin/multi-llh [options] --config <json config file>  -o <output directory> --iterations <number of iterations> 
Basic Command Line Parameter App
Options:
  -h [ --help ]                    produce help message
  -o [ --outdir ] arg (=./sample/) Directory of output
  --nsideout arg (=64)             Healpix Nside for output map
  --timesteps arg (=360)           Number of time steps
  --timestepmin arg (=0)           First time step to use
  --timestepmax arg (=0)           Last time step to use
  --iterations arg (=20)           Number of iterations
  --save-iter                      save each iteration
  -f [ --fluctuate ]               add random fluctuations
  --seed arg (=123)                RNG seed
  --iso                            make isotropic map
  --smoothing-radius arg (=1)      Smoothing radius (deg) for significance map
  --config arg (=config.json)      JSON config

Input files: The input should be specified in the JSON config file (see examples) and consists of data binned into local HEALPix maps for each sidereal time bin (24h/360 bins by default). The config file specifies the prefix and suffix of each file. For eaxmple:

  "prefix": "CR_ICECUBE_LOCAL_NSIDE64_degbin-",
  "suffix": ".fits.gz"

would assume that the files are named

  CR_ICECUBE_LOCAL_NSIDE64_degbin-000.fits.gz
  CR_ICECUBE_LOCAL_NSIDE64_degbin-001.fits.gz
  ...
  CR_ICECUBE_LOCAL_NSIDE64_degbin-359.fits.gz

Local coordinates:

The rotation matrices implemented in this code use a conversion of J2000.0 equatorial coordinates to local coordinates.

Local coordinates are expressed in a right-handed sense, with $x=E$ and $y=N$, with $\phi$ = 0 oriented along the x-axis and following the physics convention for a unit sphere with coordinates $(\theta,\phi)$, with polar angle $\theta$ (theta) (angle with respect to positive polar axis), and azimuthal angle $\phi$ (phi).

Output files: The output files will be named:

• "CR_{detector(s)}__64_360_iteration{iter}.fits.gz": (data $d_i$, background $b_i$, $a_{\ell m}$-smoothed relative intensity)
• "variance_{detector(s)}_{nside}_{nTbins}_iteration{iter}.fits.gz": (relative error $\sigma_i/\mu_i$, signed Li-Ma significance $\sigma_{i,\mathrm{Li-Ma}}$ , expectation $\mu$)
• "llhratio_{detector(s)}_{nside}_{nTbins}_iteration{iter}.dat": ($\mathrm{llh}_n - \mathrm{llh}_0$)
• "exposure_{detector}_{nside}_{nTbins}_iteration{iter}.fits.gz": ($A_i$: one file per dectector)
• "norm_{detector}_{nside}_{nTbins}_iteration{iter}.dat": ($N_\tau$: one file per detector)

If --save-iter option is used, there will be one of each of the above files per iteration. Otherwise only the initial and final results will be saved.

The code will terminate after convergence or after reaching the number of iterations (whichever comes first).