.gitlab-ci.yml

@@ -42,6 +42,9 @@ containers:
     # Fail if differences exist. If this fails, you may need to run
     # write_dockerfiles.py and commit the changes.
     - git diff --exit-code
+    - cp env-template.yml env.yml
+    - echo "  - python=3.10" >> env.yml
+    - mamba env create -f env.yml -n test --dry-run
 
 .test-python: &test-python
   stage: initial
@@ -230,6 +233,7 @@ docs:
   script:
     # Make the documentation
     - python -m pip install .
+    - python -m pip install myst_parser # only for testing purposes - remove once test image is generating correctly
     - cd examples/tutorials
     - jupyter nbconvert --to notebook --execute *.ipynb --output-dir ../../docs
     - cd ../../docs

AUTHORS.md

@@ -29,6 +29,7 @@ Gregory Ashton
 Hank Hua
 Hector Estelles
 Ignacio Magaña Hernandez
+Isaac McMahon
 Isobel Marguarethe Romero-Shaw
 Jack Heinzel
 Jacob Golomb

CHANGELOG.md

 # All notable changes will be documented in this file
 
-## [2.2.3] 2024-02-24
+## [Unreleased]
+
+
+## [2.3.0] - 2024-05-30
+
+### Added
+
+- Add support for sampler plugins via entry points (!1340, !1355)
+- Add `bilby.core.sampler.get_implemented_samplers` and `bilby.core.get_sampler_class` (!1340)
+- Add `bilby.core.utils.entry_points.get_entry_points` (!1340)
+- Add support for reading results from PathLike objects (!1342)
+- Add `snrs_as_sample` property to `bilby.gw.likelihood.base.GravitationalWaveTransient` (!1344)
+- Add `get_expected_outputs` method to the sampler classes (!1336)
+
+### Changed
+
+- Change `bilby_mcmc` to use `glasflow` instead of `nflows` (!1332)
+- Sampler classes in are no longer imported in `bilby.core.sampler` (!1340)
+- Sampler classes in `bilby.core.sampler.IMPLEMENTED_SAMPLERS` must now be loaded before use (!1340)
+- `bilby.core.sampler.IMPLEMENTED_SAMPLERS` is now an instance of `bilby.core.sampler.ImplementedSampler` instead of a dictionary (!1355)
+- Updates to support numpy v2 (!1362)
+
+### Fixed
+
+- Include final frequency point in relative binning integration (!1310)
+- Address various deprecation warnings and deprecated keyword arguments (!1316, !1326, !1343)
+- Fix typo in logging statement in `bilby.gw.source` (!1325)
+- Fix missing import in `bilby.gw.detector.load_data_from_cache_file` (!1327)
+- Fix bug where `linestyle` was ignored in `bilby.core.result.plot_multiple` (!1238)
+- Fix `soft_init` sampler keyword argument with `dynesty` (!1335)
+- Fix ZeroDivisionError when using the `dynesty` with `act-walk` and large values of `nact` (!1346)
+- Fix custom prior loading from result file (!1360)
+
+
+## [2.2.3] - 2024-02-24
 Version 2.2.3 release of Bilby
 
 This is a bugfix release 
@@ -13,7 +47,7 @@ There are also a number of testing/infrastructure updates.
 - Add the ability to change the pool size when resuming a `dynesty` job (!1315)
 - Fix how the random seed is passed to `dynesty` (!1319)
 
-## [2.2.2] 2023-11-29
+## [2.2.2] - 2023-11-29
 Version 2.2.2 release of Bilby
 
 This is a bugfix release reverting a change from 2.2.1
@@ -21,7 +55,7 @@ This is a bugfix release reverting a change from 2.2.1
 ### Changes
 - Revert !1284 (!1306)
 
-## [2.2.1] 2023-1111
+## [2.2.1] - 2023-1111
 Version 2.2.1 release of Bilby
 
 This release is a bugfix release.
@@ -36,7 +70,7 @@ This release is a bugfix release.
 - A bug with saving lists that contain None (!1301)
 - Preparatory fix an upcoming change in dynesty (!1302)
 
-## [2.2.0] 2023-07-24
+## [2.2.0] - 2023-07-24
 Version 2.2.0 release of Bilby
 
 This release contains one new feature and drops support for Python 3.8.
@@ -56,7 +90,7 @@ This release contains one new feature and drops support for Python 3.8.
 ### Deprecated
 - Drop support for py38 (!1277)
 
-## [2.1.2] 2023-07-17
+## [2.1.2] - 2023-07-17
 Version 2.1.2 release of Bilby
 
 This is a bugfix release.
@@ -70,7 +104,7 @@ Where users have previously used `np.random.seed` they should now call
 - Enable cosmological priors to be written/read in our plain text format (!1258)
 - Allow posterior reweighting to be performed when changing the likelihood and the prior (!1260)
 
-## [2.1.1] 2023-04-28
+## [2.1.1] - 2023-04-28
 Version 2.1.1 release of Bilby
 
 Bugfix release
@@ -80,7 +114,7 @@ Bugfix release
 - Bugfix for Fisher matrix proposals in `bilby_mcmc` (!1251)
 - Make the changes to the spline calibration backward compatible, 2.0.2 resume files can't be read with 2.1.0 (!1250)
 
-## [2.1.0] 2023-04-12
+## [2.1.0] - 2023-04-12
 Version 2.1.0 release of Bilby
 
 Minor feature improvements and bug fixes
@@ -103,7 +137,7 @@ Minor feature improvements and bug fixes
 ### Deprecated
 - Reading/writing ROQ weights to json (!1232)
 
-## [2.0.2] 2023-03-21
+## [2.0.2] - 2023-03-21
 Version 2.0.2 release of Bilby
 
 This is a bugfix release after the last major update.
@@ -113,7 +147,7 @@ This is a bugfix release after the last major update.
 - Fix to time calibration (!1234)
 - Fix nessai sampling time (!1236)
 
-## [2.0.1] 2023-03-13
+## [2.0.1] - 2023-03-13
 Version 2.0.1 release of Bilby
 
 This is a bugfix release after the last major update.
@@ -126,7 +160,7 @@ Users may notice changes in inferred binary neutron star masses after updating t
 - Update value for the solar mass (!1229).
 - Make `scikit-learn` an explicit dependence of `bilby[GW]` (!1230).
 
-## [2.0.0] 2023-02-29
+## [2.0.0] - 2023-02-29
 Version 2.0.0 release of Bilby
 
 This major version release has a significant change to the behaviour of the `dynesty` wrapper.
@@ -152,7 +186,7 @@ There are also a number of bugfixes and some new features in sampling and GW uti
 - Optimize ROQ waveform and calibration calls (!1216)
 - Add different proposal distribution and MCMC length for `dynesty` (!1187, !1222)
 
-## [1.4.1] 2022-12-07
+## [1.4.1] - 2022-12-07
 Version 1.4.1 release of Bilby
 
 This is a bugfix release to address some minor issues identified after v1.4.0.
@@ -166,7 +200,7 @@ This is a bugfix release to address some minor issues identified after v1.4.0.
 - Make sure that all dumping pickle files is done safely (!1189)
 - Make error catching for `dynesty` checkpointing more robust (!1190)
 
-## [1.4.0] 2022-11-18
+## [1.4.0] - 2022-11-18
 Version 1.4.0 release of Bilby
 
 The main changes in this release are support for more recent versions of `dynesty` (!1138)
@@ -188,7 +222,7 @@ and `nessai` (!1161) and adding the
 - Allow prior arguments read from a string to be functions (!1144)
 - Support `dynesty>=1.1.0` (!1138)
 
-## [1.3.0] 2022-10-23
+## [1.3.0] - 2022-10-23
 Version 1.3.0 release of Bilby
 
 This release has a major change to a sampler interface, `pymc3` is no longer supported, users should switch to `pymc>=4`.
@@ -212,7 +246,7 @@ This release also contains various documentation improvements.
 - Fix issue when specifying distance and redshfit independently (!1154)
 - Fix a bug in the storage of likelihood/prior samples for `bilby_mcmc` (!1156)
 
-## [1.2.1] 2022-09-05
+## [1.2.1] - 2022-09-05
 Version 1.2.1 release of Bilby
 
 This release contains a few bug fixes following 1.2.0.
@@ -230,7 +264,7 @@ This release contains a few bug fixes following 1.2.0.
 - Extend mass conversions to include source-frame parameters (!1131)
 - Fix prior ranges for GW150914 example (!1129)
 
-## [1.2.0] 2022-08-15
+## [1.2.0] - 2022-08-15
 Version 1.2.0 release of Bilby
 
 This is the first release that drops support for `Python<3.8`.
@@ -261,7 +295,7 @@ with multiprocessing.
 - `bilby.core.utils.progress`
 - Deepdish IO for `Result`, `Interferometer`, and `InterferometerList`
 
-## [1.1.5] 2022-01-14
+## [1.1.5] - 2022-01-14
 Version 1.1.5 release of Bilby
 
 ### Added
@@ -279,7 +313,7 @@ Version 1.1.5 release of Bilby
 - Improvements to the multi-banded GWT likelihood (!1026)
 - Improve meta data comparison (!1035)
 
-## [1.1.4] 2021-10-08
+## [1.1.4] - 2021-10-08
 Version 1.1.4 release of bilby
 
 ### Added
@@ -299,7 +333,7 @@ Version 1.1.4 release of bilby
 - Typo fix in eart light crossing (!1003)
 - Fix zero spin conversion (!1002)
 
-## [1.1.3] 2021-07-02
+## [1.1.3] - 2021-07-02
 Version 1.1.3 release of bilby
 
 ### Added
@@ -329,7 +363,7 @@ Version 1.1.3 release of bilby
 - Restructured utils module into several submodules. API remains backwards compatible (!873)
 - Changed number of default walks in `dynesty` from `10*self.ndim` to `100` (!961)
 
-## [1.1.2] 2021-05-05
+## [1.1.2] - 2021-05-05
 Version 1.1.2 release of bilby
 
 ### Added
@@ -359,13 +393,13 @@ Version 1.1.2 release of bilby
 - Fixed issues with pickle saving and loading (!932)
 - Fixed an issue with the `_base_roq_waveform` (!959)
 
-## [1.1.1] 2021-03-16
+## [1.1.1] - 2021-03-16
 Version 1.1.1 release of bilby
 
 ### Changes
 - Added `include requirements.txt` in `MANIFEST.in` to stop the pip installation from breaking
 
-## [1.1.0] 2021-03-15
+## [1.1.0] - 2021-03-15
 Version 1.1.0 release of bilby
 
 ### Added
@@ -404,7 +438,7 @@ Version 1.1.0 release of bilby
 - Fixed the likelihood count in `dynesty` (!853)
 - Changed the ordering of keyword arguments for the `Sine` and `Cosine` constructors (!892)
 
-## [1.0.4] 2020-11-23
+## [1.0.4] - 2020-11-23
 Version 1.0.4 release of bilby
 
 ### Added
@@ -413,7 +447,7 @@ Version 1.0.4 release of bilby
 ### Changes
 - Fixed issue in the CI
 
-## [1.0.3] 2020-10-23
+## [1.0.3] - 2020-10-23
 
 Version 1.0.3 release of bilby
 
@@ -432,7 +466,7 @@ Version 1.0.3 release of bilby
 - Typo fixes (!878, !887, !879)
 - Minor bug fixes (!888)
 
-## [1.0.2] 2020-09-14
+## [1.0.2] - 2020-09-14
 
 Version 1.0.2 release of bilby
 
@@ -454,7 +488,7 @@ Version 1.0.2 release of bilby
 - Clean up of code (!854)
 - Various minor bug, test and plotting fixes (!859, !874, !872, !865)
 
-## [1.0.1] 2020-08-29
+## [1.0.1] - 2020-08-29
 
 Version 1.0.1 release of bilby
 
@@ -479,7 +513,7 @@ Version 1.0.1 release of bilby
 - Various minor bug fixes and improvements to the documentation (!820)(!823)(!837)
 - Various testing improvements (!833)(!847)(!855)(!852)
 
-## [1.0.0] 2020-07-06
+## [1.0.0] - 2020-07-06
 
 Version 1.0 release of bilby
 
@@ -1030,3 +1064,33 @@ First `pip` installable version https://pypi.org/project/BILBY/ .
 
 ### Removed
 - All chainconsumer dependency as this was causing issues.
+
+
+[Unreleased]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.3.0...master
+[2.3.0]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.2.3...v2.3.0
+[2.2.3]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.2.2...v2.2.3
+[2.2.2]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.2.1...v2.2.2
+[2.2.1]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.2.0...v2.2.1
+[2.2.0]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.1.2...v2.2.0
+[2.1.2]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.1.1...v2.1.2
+[2.1.1]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.1.0...v2.1.1
+[2.1.0]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.0.2...v2.1.0
+[2.0.2]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.0.1...v2.0.2
+[2.0.1]: https://git.ligo.org/lscsoft/bilby/-/compare/v2.0.0...v2.0.1
+[2.0.0]: https://git.ligo.org/lscsoft/bilby/-/compare/v1.4.1...v2.0.0
+[1.4.1]: https://git.ligo.org/lscsoft/bilby/-/compare/v1.4.0...v1.4.1
+[1.4.0]: https://git.ligo.org/lscsoft/bilby/-/compare/1.3.0...v1.4.0
+[1.3.0]: https://git.ligo.org/lscsoft/bilby/-/compare/1.2.1...1.3.0
+[1.2.1]: https://git.ligo.org/lscsoft/bilby/-/compare/1.2.0...1.2.1
+[1.2.0]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.5...1.2.0
+[1.1.5]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.4...1.1.5
+[1.1.4]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.2...1.1.4
+[1.1.3]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.2...1.1.3
+[1.1.2]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.1...1.1.2
+[1.1.1]: https://git.ligo.org/lscsoft/bilby/-/compare/1.1.0...1.1.1
+[1.1.0]: https://git.ligo.org/lscsoft/bilby/-/compare/1.0.4...1.1.0
+[1.0.4]: https://git.ligo.org/lscsoft/bilby/-/compare/1.0.3...1.0.4
+[1.0.3]: https://git.ligo.org/lscsoft/bilby/-/compare/1.0.2...1.0.3
+[1.0.2]: https://git.ligo.org/lscsoft/bilby/-/compare/1.0.1...1.0.2
+[1.0.1]: https://git.ligo.org/lscsoft/bilby/-/compare/1.0.0...1.0.1
+[1.0.0]: https://git.ligo.org/lscsoft/bilby/-/compare/0.6.9...1.0.0

CONTRIBUTING.md

 # Contributing to bilby
 
-This is a short guide to contributing to bilby aimed at general LVC members who
+This is a short guide to contributing to bilby aimed at general LVK members who
 have some familiarity with python and git.  
 
 1. [Code of conduct](#code-of-conduct)
 2. [Code style](#code-style)
-3. [Code relevance](#code-relevance)
-4. [Merge requests](#merge-requests)
-5. [Typical workflow](#typical-workflow)
-6. [Hints and tips](#hints-and-tips)
-7. [Code overview](#code-overview)
+3. [Automated Code Checking](#automated-code-checking)
+4. [Unit Testing](#unit-testing)
+5. [Code relevance](#code-relevance)
+6. [Merge requests](#merge-requests)
+7. [Typical workflow](#typical-workflow)
+8. [Hints and tips](#hints-and-tips)
+9. [Code overview](#code-overview)
 
 
 ## Code of Conduct
@@ -17,7 +19,7 @@ have some familiarity with python and git.
 Everyone participating in the bilby community, and in particular in our issue
 tracker, merge requests, and chat channels, is expected to treat other people
 with respect and follow the guidelines articulated in the [Python Community
-Code of Conduct](https://www.python.org/psf/codeofconduct/).
+Code of Conduct](https://www.python.org/psf/codeofconduct/). Furthermore, members of the LVK collaboration must follow the [LVK Code of Conduct](https://dcc.ligo.org/LIGO-M1900037/public).
 
 ## Code style
 
@@ -46,8 +48,8 @@ def my_new_function(x, y, print=False):
 ```
 3. Avoid inline comments unless necessary. Ideally, the code should make it obvious what is going on, if not the docstring, only in subtle cases use comments
 4. Name variables sensibly. Avoid using single-letter variables, it is better to name something `power_spectral_density_array` than `psda`.
-5. Don't repeat yourself. If code is repeated in multiple places, wrap it up into a function.
-6. Add tests. The C.I. is there to do the work of "checking" the code, both now and into the future. Use it.
+5. Don't repeat yourself. If code is repeated in multiple places, wrap it up into a function. This also helps with the writing of robust unit tests (see below).
+
 
 ## Automated code checking
 
@@ -76,6 +78,50 @@ If you experience any issues with pre-commit, please ask for support on the
 usual help channels.
 
 
+## Unit Testing
+
+Unit tests are an important part of code development, helping to minimize the number of undetected bugs which may be present in a merge request. They also greatly expedite the review of code, and can even help during the initial development if used properly. Accordingly, bilby requires unit testing for any changes with machine readable inputs and outputs (i.e. pretty much everything except plotting). 
+
+Unit testing is integrated into the CI/CD pipeline, and uses the builtin unittest package. Tests should be written into the `test/` directory which corresponds to their location within the package, such that, for example, a change to `bilby/gw/conversion.py` should go into `test/gw/conversion_test.py`. To run a single test locally, one may simply do `pytest /path/to/test TestClass.test_name`, whereas to run all the tests in a given test file one may omit the class and function.
+
+For an example of what a test looks like, consider this test for the fft utils in bilby:
+
+```
+class TestFFT(unittest.TestCase):
+    def setUp(self):
+        self.sampling_frequency = 10
+
+    def tearDown(self):
+        del self.sampling_frequency
+
+    def test_nfft_sine_function(self):
+        injected_frequency = 2.7324
+        duration = 100
+        times = utils.create_time_series(self.sampling_frequency, duration)
+
+        time_domain_strain = np.sin(2 * np.pi * times * injected_frequency + 0.4)
+
+        frequency_domain_strain, frequencies = bilby.core.utils.nfft(
+            time_domain_strain, self.sampling_frequency
+        )
+        frequency_at_peak = frequencies[np.argmax(np.abs(frequency_domain_strain))]
+        self.assertAlmostEqual(injected_frequency, frequency_at_peak, places=1)
+
+    def test_nfft_infft(self):
+        time_domain_strain = np.random.normal(0, 1, 10)
+        frequency_domain_strain, _ = bilby.core.utils.nfft(
+            time_domain_strain, self.sampling_frequency
+        )
+        new_time_domain_strain = bilby.core.utils.infft(
+            frequency_domain_strain, self.sampling_frequency
+        )
+        self.assertTrue(np.allclose(time_domain_strain, new_time_domain_strain))
+```
+
+`setUp` and `tearDown` handle construction and deconstruction of the test, such that each of the other test functions may be run independently, in any order. The other two functions each make an intuitive test of the functionality of and fft/ifft function: that the fft of a sine wave should be a delta function, and that an ifft should be an inverse of an fft. 
+
+For more information on how to write effective tests, see [this guide](https://docs.python-guide.org/writing/tests/), and many others.
+
 ## Code relevance
 
 The bilby code base is intended to be highly modular and flexible. We encourage
@@ -153,7 +199,19 @@ $ git clone git@git.ligo.org:albert.einstein/bilby.git
 ```
 
 replacing the SSH url to that of your fork. This will create a directory
-`/bilby` containing a local copy of the code. From this directory, you can run
+`/bilby` containing a local copy of the code.
+
+It is strongly advised to perform development with a dedicated conda environment.
+In depth instructions for creating a conda environment may be found at the relevant
+[conda docs](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#activating-an-environment),
+but for most purposes the commands
+
+```bash
+$ conda create -n my-environment-name python=3.X
+$ conda activate my-environment-name
+```
+
+will produce the desired results. Once this is completed, one may proceed to the `/bilby` directory and run
 
 ```bash
 $ pip install -e .
@@ -202,8 +260,25 @@ $ git checkout -b my-new-feature lscsoft/master
 ### Step d) Hack away
 
 1. Develop the changes you would like to introduce, using `git add` to add files with changes. Ideally commit small units of change often, rather than creating one large commit at the end, this will simplify review and make modifying any changes easier.
-2. Commit the changes using `git commit`. This will prompt you for a commit message. Commit messages should be clear, identifying which code was changed, and why. Common practice (see e.g. [this blog](https://chris.beams.io/posts/git-commit/)) is to use a short summary line (<50 characters), followed by a blank line, then more information in longer lines.
-3. Push your changes to the remote copy of your fork on git.ligo.org
+2. Commit the changes using `git commit`. This will prompt you for a commit message. Commit messages should be clear, identifying which code was changed, and why. Bilby is adopting the use of scipy commit format, specified [here](https://docs.scipy.org/doc/scipy/dev/contributor/development_workflow.html#writing-the-commit-message). Commit messages take a standard format of `ACRONYM: Summary message` followed by a more detailed description. For example, an enhancement would look like:
+
+```
+ENH: Add my awesome new feature
+
+This is a very cool change that makes parameter estimation run 10x faster by changing a single line.
+```
+
+Similarly a bugfix:
+
+```
+BUG: Fix type error in my_awesome_feature.py
+
+Correct a typo at L1 of /bilby/my_awesome_feature.py which returned a dictionary instead of a string.
+```
+
+For more discussion of best practices, see e.g. [this blog](https://chris.beams.io/posts/git-commit/).
+
+4. Push your changes to the remote copy of your fork on git.ligo.org
 
 ```bash
 git push origin my-new-feature

README.rst

@@ -30,47 +30,7 @@ us directly. For advice on contributing, see `the contributing guide <https://gi
 Citation guide
 --------------
 
-If you use :code:`bilby` in a scientific publication, please cite
-
-* `Bilby: A user-friendly Bayesian inference library for gravitational-wave
-  astronomy
-  <https://ui.adsabs.harvard.edu/#abs/2018arXiv181102042A/abstract>`__
-* `Bayesian inference for compact binary coalescences with BILBY: validation and application to the first LIGO-Virgo gravitational-wave transient catalogue <https://ui.adsabs.harvard.edu/abs/2020MNRAS.499.3295R/abstract>`__
-
-The first of these papers introduces the software, while the second introduces advances in the sampling approaches and validation of the software.
-If you use the :code:`bilby_mcmc` sampler, please additionally cite
-
-* `BILBY-MCMC: an MCMC sampler for gravitational-wave inference <https://ui.adsabs.harvard.edu/abs/2021MNRAS.507.2037A/abstract>`__
-
-Additionally, :code:`bilby` builds on a number of open-source packages. If you
-make use of this functionality in your publications, we recommend you cite them
-as requested in their associated documentation.
-
-**Samplers**
-
-* `dynesty <https://github.com/joshspeagle/dynesty>`__
-* `nestle <https://github.com/kbarbary/nestle>`__
-* `pymultinest <https://github.com/JohannesBuchner/PyMultiNest>`__
-* `cpnest <https://github.com/johnveitch/cpnest>`__
-* `emcee <https://github.com/dfm/emcee>`__
-* `nessai <https://github.com/mj-will/nessai>`_
-* `ptemcee <https://github.com/willvousden/ptemcee>`__
-* `ptmcmcsampler <https://github.com/jellis18/PTMCMCSampler>`__
-* `pypolychord <https://github.com/PolyChord/PolyChordLite>`__
-* `PyMC3 <https://github.com/pymc-devs/pymc3>`_
-
-
-**Gravitational-wave tools**
-
-* `gwpy <https://github.com/gwpy/gwpy>`__
-* `lalsuite <https://git.ligo.org/lscsoft/lalsuite>`__
-* `astropy <https://github.com/astropy/astropy>`__
-
-**Plotting**
-
-* `corner <https://github.com/dfm/corner.py>`__ for generating corner plot
-* `matplotlib <https://github.com/matplotlib/matplotlib>`__ for general plotting routines
-
+Please refer to the `Acknowledging/citing bilby guide <https://lscsoft.docs.ligo.org/bilby/citing-bilby.html>`__.
 
 .. |pipeline status| image:: https://git.ligo.org/lscsoft/bilby/badges/master/pipeline.svg
    :target: https://git.ligo.org/lscsoft/bilby/commits/master

bilby/bilby_mcmc/sampler.py

@@ -127,6 +127,9 @@ class Bilby_MCMC(MCMCSampler):
     initial_sample_dict: dict
         A dictionary of the initial sample value. If incomplete, will overwrite
         the initial_sample drawn using initial_sample_method.
+    normalize_prior: bool
+        When False, disables calculation of constraint normalization factor
+        during prior probability computation. Default value is True.
     verbose: bool
         Whether to print diagnostic output during the run.
 
@@ -175,6 +178,7 @@ class Bilby_MCMC(MCMCSampler):
         resume=True,
         exit_code=130,
         verbose=True,
+        normalize_prior=True,
         **kwargs,
     ):
 
@@ -194,6 +198,7 @@ class Bilby_MCMC(MCMCSampler):
         self.kwargs["target_nsamples"] = self.kwargs["nsamples"]
         self.L1steps = self.kwargs["L1steps"]
         self.L2steps = self.kwargs["L2steps"]
+        self.normalize_prior = normalize_prior
         self.pt_inputs = ParallelTemperingInputs(
             **{key: self.kwargs[key] for key in ParallelTemperingInputs._fields}
         )
@@ -309,6 +314,7 @@ class Bilby_MCMC(MCMCSampler):
             evidence_method=self.evidence_method,
             initial_sample_method=self.initial_sample_method,
             initial_sample_dict=self.initial_sample_dict,
+            normalize_prior=self.normalize_prior,
         )
 
     def get_setup_string(self):
@@ -382,7 +388,9 @@ class Bilby_MCMC(MCMCSampler):
             If true, resume file was successfully loaded, otherwise false
 
         """
-        if os.path.isfile(self.resume_file) is False:
+        if os.path.isfile(self.resume_file) is False or not os.path.getsize(
+            self.resume_file
+        ):
             return False
         import dill
 
@@ -547,6 +555,29 @@ class Bilby_MCMC(MCMCSampler):
                         all_samples=ptsampler.samples,
                     )
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            The label for the run.
+
+        Returns
+        -------
+        list
+            List of file names.
+        list
+            List of directory names. Will always be empty for bilby_mcmc.
+        """
+        filenames = [os.path.join(outdir, f"{label}_resume.pickle")]
+        return filenames, []
+
 
 class BilbyPTMCMCSampler(object):
     def __init__(
@@ -560,11 +591,13 @@ class BilbyPTMCMCSampler(object):
         evidence_method,
         initial_sample_method,
         initial_sample_dict,
+        normalize_prior=True,
     ):
         self.set_pt_inputs(pt_inputs)
         self.use_ratio = use_ratio
         self.initial_sample_method = initial_sample_method
         self.initial_sample_dict = initial_sample_dict
+        self.normalize_prior = normalize_prior
         self.setup_sampler_dictionary(convergence_inputs, proposal_cycle)
         self.set_convergence_inputs(convergence_inputs)
         self.pt_rejection_sample = pt_rejection_sample
@@ -635,6 +668,7 @@ class BilbyPTMCMCSampler(object):
                     use_ratio=self.use_ratio,
                     initial_sample_method=self.initial_sample_method,
                     initial_sample_dict=self.initial_sample_dict,
+                    normalize_prior=self.normalize_prior,
                 )
                 for Eindex in range(n)
             ]
@@ -1129,12 +1163,13 @@ class BilbyMCMCSampler(object):
         use_ratio=False,
         initial_sample_method="prior",
         initial_sample_dict=None,
+        normalize_prior=True,
     ):
         self.beta = beta
         self.Tindex = Tindex
         self.Eindex = Eindex
         self.use_ratio = use_ratio
-
+        self.normalize_prior = normalize_prior
         self.parameters = _sampling_convenience_dump.priors.non_fixed_keys
         self.ndim = len(self.parameters)
 
@@ -1209,7 +1244,10 @@ class BilbyMCMCSampler(object):
         return logl
 
     def log_prior(self, sample):
-        return _sampling_convenience_dump.priors.ln_prob(sample.parameter_only_dict)
+        return _sampling_convenience_dump.priors.ln_prob(
+            sample.parameter_only_dict,
+            normalized=self.normalize_prior,
+        )
 
     def accept_proposal(self, prop, proposal):
         self.chain.append(prop)

bilby/core/prior/dict.py

@@ -537,7 +537,7 @@ class PriorDict(dict):
                 constrained_prob[keep] = prob[keep] * ratio
                 return constrained_prob
 
-    def ln_prob(self, sample, axis=None):
+    def ln_prob(self, sample, axis=None, normalized=True):
         """
 
         Parameters
@@ -546,6 +546,9 @@ class PriorDict(dict):
             Dictionary of the samples of which to calculate the log probability
         axis: None or int
             Axis along which the summation is performed
+        normalized: bool
+            When False, disables calculation of constraint normalization factor
+            during prior probability computation. Default value is True.
 
         Returns
         =======
@@ -554,10 +557,14 @@ class PriorDict(dict):
 
         """
         ln_prob = np.sum([self[key].ln_prob(sample[key]) for key in sample], axis=axis)
-        return self.check_ln_prob(sample, ln_prob)
+        return self.check_ln_prob(sample, ln_prob,
+                                  normalized=normalized)
 
-    def check_ln_prob(self, sample, ln_prob):
-        ratio = self.normalize_constraint_factor(tuple(sample.keys()))
+    def check_ln_prob(self, sample, ln_prob, normalized=True):
+        if normalized:
+            ratio = self.normalize_constraint_factor(tuple(sample.keys()))
+        else:
+            ratio = 1
         if np.all(np.isinf(ln_prob)):
             return ln_prob
         else:
@@ -785,7 +792,7 @@ class ConditionalPriorDict(PriorDict):
         prob = np.prod(res, **kwargs)
         return self.check_prob(sample, prob)
 
-    def ln_prob(self, sample, axis=None):
+    def ln_prob(self, sample, axis=None, normalized=True):
         """
 
         Parameters
@@ -794,6 +801,9 @@ class ConditionalPriorDict(PriorDict):
             Dictionary of the samples of which we want to have the log probability of
         axis: Union[None, int]
             Axis along which the summation is performed
+        normalized: bool
+            When False, disables calculation of constraint normalization factor
+            during prior probability computation. Default value is True.
 
         Returns
         =======
@@ -806,7 +816,8 @@ class ConditionalPriorDict(PriorDict):
             for key in sample
         ]
         ln_prob = np.sum(res, axis=axis)
-        return self.check_ln_prob(sample, ln_prob)
+        return self.check_ln_prob(sample, ln_prob,
+                                  normalized=normalized)
 
     def cdf(self, sample):
         self._prepare_evaluation(*zip(*sample.items()))

bilby/core/sampler/__init__.py

@@ -13,7 +13,74 @@ from ..utils import (
 from . import proposal
 from .base_sampler import Sampler, SamplingMarginalisedParameterError
 
-IMPLEMENTED_SAMPLERS = get_entry_points("bilby.samplers")
+
+class ImplementedSamplers:
+    """Dictionary-like object that contains implemented samplers.
+
+    This class is singleton and only one instance can exist.
+    """
+
+    _instance = None
+
+    _samplers = get_entry_points("bilby.samplers")
+
+    def keys(self):
+        """Iterator of available samplers by name.
+
+        Reduces the list to its simplest. This includes removing the 'bilby.'
+        prefix from native samplers if a corresponding plugin is not available.
+        """
+        keys = []
+        for key in self._samplers.keys():
+            name = key.replace("bilby.", "")
+            if name in self._samplers.keys():
+                keys.append(key)
+            else:
+                keys.append(name)
+        return iter(keys)
+
+    def values(self):
+        """Iterator of sampler classes.
+
+        Note: the classes need to loaded using :code:`.load()` before being
+        called.
+        """
+        return iter(self._samplers.values())
+
+    def items(self):
+        """Iterator of tuples containing keys (sampler names) and classes.
+
+        Note: the classes need to loaded using :code:`.load()` before being
+        called.
+        """
+        return iter(((k, v) for k, v in zip(self.keys(), self.values())))
+
+    def valid_keys(self):
+        """All valid keys including bilby.<sampler name>."""
+        keys = set(self._samplers.keys())
+        return iter(keys.union({k.replace("bilby.", "") for k in keys}))
+
+    def __getitem__(self, key):
+        if key in self._samplers:
+            return self._samplers[key]
+        elif f"bilby.{key}" in self._samplers:
+            return self._samplers[f"bilby.{key}"]
+        else:
+            raise ValueError(
+                f"Sampler {key} is not implemented! "
+                f"Available samplers are: {list(self.keys())}"
+            )
+
+    def __contains__(self, value):
+        return value in self.valid_keys()
+
+    def __new__(cls):
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+        return cls._instance
+
+
+IMPLEMENTED_SAMPLERS = ImplementedSamplers()
 
 
 def get_implemented_samplers():
@@ -51,14 +118,7 @@ def get_sampler_class(sampler):
     ValueError
         Raised if the sampler is not implemented.
     """
-    sampler = sampler.lower()
-    if sampler in IMPLEMENTED_SAMPLERS:
-        return IMPLEMENTED_SAMPLERS[sampler].load()
-    else:
-        raise ValueError(
-            f"Sampler {sampler} not yet implemented. "
-            f"The available samplers are: {get_implemented_samplers()}"
-        )
+    return IMPLEMENTED_SAMPLERS[sampler.lower()].load()
 
 
 if command_line_args.sampler_help:

bilby/core/sampler/base_sampler.py

@@ -173,6 +173,13 @@ class Sampler(object):
         Whether the implemented sampler exits hard (:code:`os._exit` rather
         than :code:`sys.exit`). The latter can be escaped as :code:`SystemExit`.
         The former cannot.
+    sampler_name : str
+        Name of the sampler. This is used when creating the output directory for
+        the sampler.
+    abbreviation : str
+        Abbreviated name of the sampler. Does not have to be specified in child
+        classes. If set to a value other than :code:`None`, this will be used
+        instead of :code:`sampler_name` when creating the output directory.
 
     Raises
     ======
@@ -187,6 +194,8 @@ class Sampler(object):
 
     """
 
+    sampler_name = "sampler"
+    abbreviation = None
     default_kwargs = dict()
     npool_equiv_kwargs = [
         "npool",
@@ -454,7 +463,7 @@ class Sampler(object):
             logger.info("Unable to measure single likelihood time")
         else:
             logger.info(
-                f"Single likelihood evaluation took {self._log_likelihood_eval_time:.3e} s"
+                f"Single likelihood evaluation took {log_likelihood_eval_time:.3e} s"
             )
         return log_likelihood_eval_time
 
@@ -779,8 +788,37 @@ class Sampler(object):
     def write_current_state(self):
         raise NotImplementedError()
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+        Both can be empty. Defaults to a single directory:
+        :code:`"{outdir}/{name}_{label}/"`, where :code:`name`
+        is :code:`abbreviation` if it is defined for the sampler class, otherwise
+        it defaults to :code:`sampler_name`.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            The label for the run.
+
+        Returns
+        -------
+        list
+            List of file names.
+        list
+            List of directory names.
+        """
+        name = cls.abbreviation or cls.sampler_name
+        dirname = os.path.join(outdir, f"{name}_{label}", "")
+        return [], [dirname]
+
 
 class NestedSampler(Sampler):
+    sampler_name = "nested_sampler"
     npoints_equiv_kwargs = [
         "nlive",
         "nlives",
@@ -854,6 +892,7 @@ class NestedSampler(Sampler):
 
 
 class MCMCSampler(Sampler):
+    sampler_name = "mcmc_sampler"
     nwalkers_equiv_kwargs = ["nwalker", "nwalkers", "draws", "Niter"]
     nburn_equiv_kwargs = ["burn", "nburn"]
 

bilby/core/sampler/cpnest.py

@@ -40,6 +40,7 @@ class Cpnest(NestedSampler):
 
     """
 
+    sampler_name = "cpnest"
     default_kwargs = dict(
         verbose=3,
         nthreads=1,

bilby/core/sampler/dnest4.py

@@ -99,6 +99,7 @@ class DNest4(_TemporaryFileSamplerMixin, NestedSampler):
         If True, prints information during run
     """
 
+    sampler_name = "d4nest"
     default_kwargs = dict(
         max_num_levels=20,
         num_steps=500,

bilby/core/sampler/dynamic_dynesty.py

@@ -14,6 +14,7 @@ class DynamicDynesty(Dynesty):
     """
 
     external_sampler_name = "dynesty"
+    sampler_name = "dynamic_dynesty"
 
     @property
     def nlive(self):

bilby/core/sampler/dynesty.py

@@ -3,6 +3,7 @@ import inspect
 import os
 import sys
 import time
+import warnings
 
 import numpy as np
 from pandas import DataFrame
@@ -151,6 +152,7 @@ class Dynesty(NestedSampler):
         specified.
     """
 
+    sampler_name = "dynesty"
     sampling_seed_key = "seed"
 
     @property
@@ -299,6 +301,32 @@ class Dynesty(NestedSampler):
                 )
         Sampler._verify_kwargs_against_default_kwargs(self)
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            The label for the run.
+
+        Returns
+        -------
+        list
+            List of file names.
+        list
+            List of directory names. Will always be empty for dynesty.
+        """
+        filenames = []
+        for kind in ["resume", "dynesty"]:
+            filename = os.path.join(outdir, f"{label}_{kind}.pickle")
+            filenames.append(filename)
+        return filenames, []
+
     def _print_func(
         self,
         results,
@@ -650,12 +678,21 @@ class Dynesty(NestedSampler):
         chain of nested samples within dynesty and have to be removed before
         restarting the sampler.
         """
+
         logger.debug("Running sampler with checkpointing")
 
         old_ncall = self.sampler.ncall
         sampler_kwargs = self.sampler_function_kwargs.copy()
+        warnings.filterwarnings(
+            "ignore",
+            message="The sampling was stopped short due to maxiter/maxcall limit*",
+            category=UserWarning,
+            module="dynesty.sampler",
+        )
         while True:
             self.finalize_sampler_kwargs(sampler_kwargs)
+            if getattr(self.sampler, "added_live", False):
+                self.sampler._remove_live_points()
             self.sampler.run_nested(**sampler_kwargs)
             if self.sampler.ncall == old_ncall:
                 break
@@ -670,8 +707,8 @@ class Dynesty(NestedSampler):
             if last_checkpoint_s > self.check_point_delta_t:
                 self.write_current_state()
                 self.plot_current_state()
-            if getattr(self.sampler, "added_live", False):
-                self.sampler._remove_live_points()
+        if getattr(self.sampler, "added_live", False):
+            self.sampler._remove_live_points()
 
         self.sampler.run_nested(**sampler_kwargs)
         self.write_current_state()
@@ -709,7 +746,10 @@ class Dynesty(NestedSampler):
         if os.path.isfile(self.resume_file):
             logger.info(f"Reading resume file {self.resume_file}")
             with open(self.resume_file, "rb") as file:
-                sampler = dill.load(file)
+                try:
+                    sampler = dill.load(file)
+                except EOFError:
+                    sampler = None
 
                 if not hasattr(sampler, "versions"):
                     logger.warning(

bilby/core/sampler/emcee.py

@@ -45,6 +45,7 @@ class Emcee(MCMCSampler):
 
     """
 
+    sampler_name = "emcee"
     default_kwargs = dict(
         nwalkers=500,
         a=2,
@@ -310,7 +311,11 @@ class Emcee(MCMCSampler):
         """
         if hasattr(self, "_sampler"):
             pass
-        elif self.resume and os.path.isfile(self.checkpoint_info.sampler_file):
+        elif (
+            self.resume
+            and os.path.isfile(self.checkpoint_info.sampler_file)
+            and os.path.getsize(self.checkpoint_info.sampler_file)
+        ):
             import dill
 
             logger.info(

bilby/core/sampler/fake_sampler.py

@@ -17,6 +17,8 @@ class FakeSampler(Sampler):
         A string pointing to the posterior data file to be loaded.
     """
 
+    sampler_name = "fake_sampler"
+
     default_kwargs = dict(
         verbose=True, logl_args=None, logl_kwargs=None, print_progress=True
     )

bilby/core/sampler/kombine.py

@@ -39,6 +39,7 @@ class Kombine(Emcee):
 
     """
 
+    sampler_name = "kombine"
     default_kwargs = dict(
         nwalkers=500,
         args=[],
@@ -165,7 +166,11 @@ class Kombine(Emcee):
         return self.sampler.chain[:nsteps, :, :]
 
     def check_resume(self):
-        return self.resume and os.path.isfile(self.checkpoint_info.sampler_file)
+        return (
+            self.resume
+            and os.path.isfile(self.checkpoint_info.sampler_file)
+            and os.path.getsize(self.checkpoint_info.sampler_file) > 0
+        )
 
     @signal_wrapper
     def run_sampler(self):

bilby/core/sampler/nessai.py

@@ -20,6 +20,7 @@ class Nessai(NestedSampler):
     Documentation: https://nessai.readthedocs.io/
     """
 
+    sampler_name = "nessai"
     _default_kwargs = None
     _run_kwargs_list = None
     sampling_seed_key = "seed"
@@ -300,5 +301,30 @@ class Nessai(NestedSampler):
         self._log_interruption(signum=signum)
         sys.exit(self.exit_code)
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            The label for the run.
+
+        Returns
+        -------
+        list
+            List of file names. This will be empty for nessai.
+        list
+            List of directory names.
+        """
+        dirs = [os.path.join(outdir, f"{label}_{cls.sampler_name}", "")]
+        dirs += [os.path.join(dirs[0], d, "") for d in ["proposal", "diagnostics"]]
+        filenames = []
+        return filenames, dirs
+
     def _setup_pool(self):
         pass

bilby/core/sampler/nestle.py

@@ -24,6 +24,7 @@ class Nestle(NestedSampler):
 
     """
 
+    sampler_name = "nestle"
     default_kwargs = dict(
         verbose=True,
         method="multi",

bilby/core/sampler/polychord.py

+import os
+
 import numpy as np
 
 from .base_sampler import NestedSampler, signal_wrapper
@@ -20,6 +22,7 @@ class PyPolyChord(NestedSampler):
     To see what the keyword arguments are for, see the docstring of PyPolyChordSettings
     """
 
+    sampler_name = "pypolychord"
     default_kwargs = dict(
         use_polychord_defaults=False,
         nlive=None,
@@ -130,6 +133,28 @@ class PyPolyChord(NestedSampler):
         physical_parameters = samples[:, -self.ndim :]
         return log_likelihoods, physical_parameters
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            Ignored for pypolychord.
+
+        Returns
+        -------
+        list
+            List of file names. This will always be empty for pypolychord.
+        list
+            List of directory names.
+        """
+        return [], [os.path.join(outdir, "chains", "")]
+
     @property
     def _sample_file_directory(self):
         return self.outdir + "/chains"

bilby/core/sampler/ptemcee.py

@@ -128,6 +128,7 @@ class Ptemcee(MCMCSampler):
 
     """
 
+    sampler_name = "ptemcee"
     # Arguments used by ptemcee
     default_kwargs = dict(
         ntemps=10,
@@ -414,7 +415,11 @@ class Ptemcee(MCMCSampler):
             # This is a very ugly hack to support numpy>=1.24
             ptemcee.sampler.np.float = float
 
-        if os.path.isfile(self.resume_file) and self.resume is True:
+        if (
+            os.path.isfile(self.resume_file)
+            and os.path.getsize(self.resume_file)
+            and self.resume is True
+        ):
             import dill
 
             logger.info(f"Resume data {self.resume_file} found")
@@ -512,7 +517,7 @@ class Ptemcee(MCMCSampler):
         logger.info("Starting to sample")
 
         while True:
-            for (pos0, log_posterior, log_likelihood) in sampler.sample(
+            for pos0, log_posterior, log_likelihood in sampler.sample(
                 self.pos0,
                 storechain=False,
                 iterations=self.convergence_inputs.niterations_per_check,
@@ -710,6 +715,29 @@ class Ptemcee(MCMCSampler):
             except Exception as e:
                 logger.info(f"mean_logl plot failed with exception {e}")
 
+    @classmethod
+    def get_expected_outputs(cls, outdir=None, label=None):
+        """Get lists of the expected outputs directories and files.
+
+        These are used by :code:`bilby_pipe` when transferring files via HTCondor.
+
+        Parameters
+        ----------
+        outdir : str
+            The output directory.
+        label : str
+            The label for the run.
+
+        Returns
+        -------
+        list
+            List of file names.
+        list
+            List of directory names. Will always be empty for ptemcee.
+        """
+        filenames = [f"{outdir}/{label}_checkpoint_resume.pickle"]
+        return filenames, []
+
 
 def get_minimum_stable_itertion(mean_array, frac, nsteps_min=10):
     nsteps = mean_array.shape[1]