Skip to content

Releases: joezuntz/cosmosis

v2.3

03 Mar 12:17
404b914
Compare
Choose a tag to compare

What's Changed

New Contributors

Full Changelog: v2.2.5...v2.3

v2.2.5

20 Feb 17:38
Compare
Choose a tag to compare

What's Changed

  • Add option to draw starting sample from proposal by @joezuntz in #63
  • Better error messages and tests for them by @joezuntz in #66
  • Fixes for MN and PC output of vector quantities by @joezuntz in #71

Full Changelog: v2.2.3...v2.2.5

v2.2.4

20 Feb 17:35
8c7a82c
Compare
Choose a tag to compare

What's Changed

  • Add option to draw starting sample from proposal by @joezuntz in #63
  • Better error messages and tests for them by @joezuntz in #66
  • Fixes for MN and PC output of vector quantities by @joezuntz in #71

Full Changelog: v2.2.3...v2.2.4

v2.2.3

08 Nov 10:10
e3e53cd
Compare
Choose a tag to compare

What's Changed

Full Changelog: v2.2.2...v2.2.3

v2.2.2

17 Oct 17:06
Compare
Choose a tag to compare

What's Changed

CosmoSIS previously launched git in a subprocess to try and find out the git revisions of various directories. This causes problems on the NERSC Perlmutter machine because that's technically undefined behaviour under MPI and the Cray MPI implementation actually does do strange things. CosmoSIS will now try to use the Dulwich library to find the versions without using a subprocess, and in either case the setting the new COSMOSIS_NO_SUBPROCESS environment variable to 1 will stop any subprocesses from being launched.

We also now test CosmoSIS under python 3.10, so that version is now fully supported.

Full Changelog: v2.2.2...v2.2

v2.2.1

17 Oct 17:03
695c258
Compare
Choose a tag to compare

What's Changed

CosmoSIS previously launched git in a subprocess to try and find out the git revisions of various directories. This causes problems on the NERSC Perlmutter machine because that's technically undefined behaviour under MPI and the Cray MPI implementation actually does do strange things. CosmoSIS will now try to use the Dulwich library to find the versions without using a subprocess, and in either case the setting the new COSMOSIS_NO_SUBPROCESS environment variable to 1 will stop any subprocesses from being launched.

We also now test CosmoSIS under python 3.10, so that version is now fully supported.

Full Changelog: v2.2...v2.2.1

v2.2

21 Sep 15:29
90504b1
Compare
Choose a tag to compare

What's Changed

  • Added a flag --profile FILENAME to cosmosis to use the python cProfile tool to see where time is spent in the code
  • Added the flag --truth FILENAME to cosmosis-postprocess which adds a truth marker when given an values file
  • Fixed a problem with contour legends in postprocessing in newer matplotlib versions

Full Changelog: v2.1.5...v2.2

v2.1.5

01 Sep 11:07
c974b35
Compare
Choose a tag to compare

What's Changed

Full Changelog: v2.1.4...v2.1.5

v2.1.4

16 Aug 15:42
01f53e4
Compare
Choose a tag to compare

What's Changed

  • Allow new parameters to be added using the -v flag by @joezuntz in #41

Full Changelog: v2.1.3...v2.1.4

v2.1.3

23 Jul 12:14
Compare
Choose a tag to compare

This version contains the same contents as 2.1.0:

  • Expose new PocoMC sampler by @joezuntz in #37
  • Generate trace plots and improve dynesty integration by @joezuntz in #38