search

joshspeagle / dynesty

Dynamic Nested Sampling package for computing Bayesian posteriors and evidences
https://dynesty.readthedocs.io/
MIT License
357 stars 77 forks source link

Ellipsoid decomposition failing (v0.9.2) #99

Closed joshspeagle closed 6 years ago

joshspeagle commented 6 years ago

Looks like the OAS method is often over-regularizing the ellipsoids, leading to poor performance. There are also some issues with singular decompositions still sometimes being returned.

joshspeagle commented 6 years ago

This is tentatively resolved in https://github.com/joshspeagle/dynesty/commit/50483817ae2de6bf24d2e711a71e900a7fa5a778, so I'm closing this unless it becomes clear that the ellipsoids are still a problem.