Welcome to Sky Area’s documentation!

The sky_area package provides utilities to turn samples from probability distributions on the sky (i.e. samples of RA and DEC) into sky maps, computing credible areas of the distribution, calculating the minimum searched area following a greedy algorithm to find an object, and producing Healpix-pixellated maps that can be used to optimise a search with a known telescope beam.

There are also executable codes, that rely on the LALInference libraries from the LIGO Scientific Collaboration, for producing various skymaps and credible regions in FITS format.

The algorithm used to turn discrete samples into a probability distribution on the sky is an improved version of the clustering algorithm X-means that provides more flexibility in the shape of each cluster. The code works hard to ensure that the quoted credible areas are unbiased, so the X% credible area will, on average, enclose X% of the probability mass.

You may want to:

  • Compute the credible areas or the area searched under a greedy algorithm for a distribution on the sky represented by discrete samples. Use the sky_area.sky_area_clustering.ClusteredSkyKDEPosterior class.
  • Automatically produce the above from the output of a LALInference run. Use the executable program run_sky_area.py
  • Produce a Healpix map that ranks pixels on the sky for a search following the posterior denisty with a telescope having a known beam size. Use the sky_area.search.search_map() function or, from the command-line, the make_search_map.py executable.
  • Collate a bunch of sky maps, searched areas, and credible areas to produce a cumulative distribution of searched/credible areas from a combined data set of posteriors, as in Singer, et al. Use the process_areas.py executable.
  • Compute, as a function of position on the sky, the constraints on the distance of the source. Use the sky_area.sky_area_clustering.Clustered3DKDEPosterior.

The sky_area.sky_area_clustering Module

The sky_area.search Module

Indices and tables