Skip to content

Constellations

Reed Essick requested to merge constellations into master

This moves around a bunch of stuff in distributions.locations. This is done to make it easier to organize where a few things are declared (eg, grouping all the western constellations in a single file that is separate from other constellations).

Fixes #19 (closed).

To Do:

  • write a special case of RADecDistribution (called Constellation) that is a top-hat for a given polygonal bounding box.
  • Test Constellation via gwdistributions-test-constellation
  • write an auto-declaration loop that iterates through constellations
    • this is actually done by loading vertices at instantiation time based on a supplied "name"
  • write a mixture model over all Constellation classes. Parameters are the probabilities of that constellation (note that larger constellations should have larger probabilities).
    • do this for separate zodiacs? all the ones you can find
      • IAU (western zodiac) : this is done
      • chinese zodiac?
      • else?
  • write a test of IAUConstellation and IAUConstellationMixture
    • a one-off script that makes mollweide plots with vertices, edges, and drawn samples
    • make sure that I'm doing something reasonable for all constellations and there is no "bad behavior" associated with constellations that wrap around the periodic boundary condition at RA=2pi
Edited by Reed Essick

Merge request reports

Loading