Bruce Edelman authored 5 years ago and Gregory Ashton committed 5 years ago

generalized the JointPrior object structure from Matthew Pitkin's MutlivariateGaussian prior formalism. TODO: add in the joint MapPrior for HEALPix priors

Bruce Edelman authored 5 years ago and Moritz Huebner committed 5 years ago

generalized the JointPrior object structure from Matthew Pitkin's MutlivariateGaussian prior formalism. TODO: add in the joint MapPrior for HEALPix priors

Gregory Ashton authored 5 years ago and Moritz Huebner committed 5 years ago

From python3.6+, all dicts are ordered by design. On the otherhand, the
collections OrderedDict has odd behaviour when pickled (see comment
below). This changes to just using dicts. For users of python3.5 and
earlier, this will only change the order of their output posterior. On
the other hand, it will make pickling priors much more straight forward
going forward.

1) Make the constants consistent with those used in lal
2) Deprecate the gps_time_to_gmst function
3) Use the lal.GreenwichMeanSiderealTime(time) method directly

- Ensures the periodic argument is passed through to the dynesty making
  the points wander outside the allowed range
- Fix tests

?

Reapplies the 10xndim default setting. Whilethe 30xndim is more robust,
since we wont use default settings to pass the PP test it is overly
burdensom of higher-mass runs to keep with this default.

Matthew David Pitkin authored 5 years ago and Colm Talbot committed 5 years ago

 - to do this a MultivariateGaussianDist must also be specified and parsed
   by the prior file reading function

This adds a test-mode command like argument to run with zero likelihood.
As such, this will sample from the priors alone. To use, run any
standard bilby example with `--bilby-zero-likelihood-mode` in the
command line. This has been intentionally obfuscated to avoid
overloading the usual documentation.