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, themake_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
.