v2.8.1

This is a bugfix release. It addresses minors bugs and fixes support for numpy 2.4.

Fixes

• BUG: Fix 0-dimensional array in priors by @ColmTalbot in #886 and backported #1103
• BUG: Fix typo in interferometer.py to_pickle method by @AlexandreGoettel in #1081 and backported in #1103

Changes

• MAINT: Relax tolerance for log noise evidence comparison by @mattpitkin in #1087 and backported in #1103

Full Changelog: v2.8.0...v2.8.1

v2.8.0

Fixes

• BUG: Fix handling of contour-kwargs for multiple corner plots by @JasperMartins in #1022
• BUG: Exclude infinite values from gradient check tau list by @ColmTalbot in #1035
• BUG: fix triangular interferometer coordinates by @filippo-santoliquido in #1016
• BUG: Fix path for LinAlgError in exception handling by @ColmTalbot in #1038
• BUG: inhomogeneous rescaling with Categorical Priors by @Hen42rik in #1037
• BUG: Fix GWSignalWaveformGenerator reconstruction in waveform posterior plots by @lorenzopompili00 in #1029
• BUG: Fix sampler_kwargs mutation in dynesty3_utils by @ColmTalbot in #1045

Changes

• MAINT: Change probability initialization to use np.ones_like by @vivienr in #1028
• REFACTOR: make multiband likelihood call Interferometer.get_detector_response by @ColmTalbot in #847
• MAINT: move LikePriorEvaluator to base sampler by @mj-will in #1041
• MAINT: Advance deprecations before bilby v3 by @ColmTalbot in #1040

New Contributors

• @Dsantra92 made their first contribution in #1013
• @vivienr made their first contribution in #1028
• @Qazalbash made their first contribution in #1000
• @filippo-santoliquido made their first contribution in #1016

Full Changelog: v2.7.1...v2.8.0

v2.7.1

This is a bugfix release addressing two bugs in using the new likelihood.log_likelihood(parameters) API and one in specifying the cosmology.

Fixes

• The global cosmology was not being set properly by @mj-will in #1005
• The noise_log_likelihood method was erroneously given a parameters argument in several likelihood classes by @Hen42rik in #1010
• The Fisher matrix posterior estimator was not passed all needed parameters in bilby_mcmc by @ColmTalbot in #1012

New Contributors

• @Hen42rik made their first contribution in #1010

Full Changelog: v2.7.0...v2.7.1

v2.7.0

There are a few major changes/additions in this release along with minor changes and removals.

Major changes

• Likelihood instances can now be called as likelihood.log_likelihood(parameters), see here for more information.
• Support the new API in dynesty=3
• Add a new WaveformGenerator capable of using arbitrary waveform models implemented through the gwsignal waveform interface.

Additions

• Added WeightedCategorical-Prior by @JasperMartins in #893
• ENH: Implement DiscreteValues prior by @unkaktus in #947
• ENH: Allow no parameters as state by @ColmTalbot in #941
• ENH: add support for new dynesty api by @ColmTalbot in #950
• FEAT: add gwsignal waveform generator by @ColmTalbot in #877

Fixes

• BUG: Fix sampling efficiency warning by @fgittins in #953
• TYPO: fix random call in example by @ColmTalbot in #973
• BUG: Fix matched filter SNR calculation in time domain injection by @ColmTalbot in #957
• TST: mark whitened strain tests as flaky by @mj-will in #987
• MAINT: update file extension logic by @mj-will in #962
• BUG: Custom prior loading by @asb5468 in #940
• BUG: Update sampler_init_kwargs for dynestyv3 compatibility by @ColmTalbot in #994
• BUG: Refactor mapper assignment for new dynesty API by @ColmTalbot in #998
• BUG: fix how kwargs are passed to custom dynesty samplers by @ColmTalbot #999

Changes

• ENH: plot_multiple: Allow plotting onto user-defined figure by @unkaktus in #946
• MAINT: change RNG imports by @mj-will in #943
• MAINT: np.trapz -> np.trapezoid by @ColmTalbot in #974
• MAINT: define Planck15-LAL cosmology using LAL constants by @mj-will in #932
• DEV: make sure all priors return float when needed by @ColmTalbot in #979
• Replace pytables with h5py by @duncanmmacleod in #982

Deprecations

• DEP: deprecate dnest4 interface by @ColmTalbot in #980

Removed

• MAINT: remove unsupported roq json weight file format by @ColmTalbot in #945

Other changes

• DOC: Fixed README link for opening bilby_pipe issues by @mick-wright in #922
• DOC: Correct docs for use_ratio argument of run_sampler by @mattpitkin in #958
• CI: add merge_group to enable merge queue by @mj-will in #983
• TST: remove dnest4 import test by @ColmTalbot in #984
• BLD: migrate to pyproject.toml by @mj-will in #952
• BLD: use release branches by @mj-will in #954

New Contributors

• @fgittins made their first contribution in #953
• @duncanmmacleod made their first contribution in #982

Full Changelog: v2.6.0...v2.7.0

v2.7.0rc1

This is a release candidate for Bilby 2.7.0. There are a few major changes/additions in this release along with minor changes and removals.

Major changes

• Likelihood instances can now be called as likelihood.log_likelihood(parameters), see here for more information.
• Support the new API in dynesty=3
• Add a new WaveformGenerator capable of using arbitrary waveform models implemented through the gwsignal waveform interface.

Additions

• Added WeightedCategorical-Prior by @JasperMartins in #893
• ENH: Implement DiscreteValues prior by @unkaktus in #947
• ENH: Allow no parameters as state by @ColmTalbot in #941
• ENH: add support for new dynesty api by @ColmTalbot in #950
• FEAT: add gwsignal waveform generator by @ColmTalbot in #877

Fixes

• BUG: Fix sampling efficiency warning by @fgittins in #953
• TYPO: fix random call in example by @ColmTalbot in #973
• BUG: Fix matched filter SNR calculation in time domain injection by @ColmTalbot in #957
• TST: mark whitened strain tests as flaky by @mj-will in #987
• MAINT: update file extension logic by @mj-will in #962
• BUG: Custom prior loading by @asb5468 in #940
• BUG: Update sampler_init_kwargs for dynestyv3 compatibility by @ColmTalbot in #994
• BUG: Refactor mapper assignment for new dynesty API by @ColmTalbot in #998

Changes

• ENH: plot_multiple: Allow plotting onto user-defined figure by @unkaktus in #946
• MAINT: change RNG imports by @mj-will in #943
• MAINT: np.trapz -> np.trapezoid by @ColmTalbot in #974
• MAINT: define Planck15-LAL cosmology using LAL constants by @mj-will in #932
• DEV: make sure all priors return float when needed by @ColmTalbot in #979
• Replace pytables with h5py by @duncanmmacleod in #982

Deprecations

• DEP: deprecate dnest4 interface by @ColmTalbot in #980

Removed

• MAINT: remove unsupported roq json weight file format by @ColmTalbot in #945

Other changes

• DOC: Fixed README link for opening bilby_pipe issues by @mick-wright in #922
• DOC: Correct docs for use_ratio argument of run_sampler by @mattpitkin in #958
• CI: add merge_group to enable merge queue by @mj-will in #983
• TST: remove dnest4 import test by @ColmTalbot in #984
• BLD: migrate to pyproject.toml by @mj-will in #952
• BLD: use release branches by @mj-will in #954

New Contributors

• @fgittins made their first contribution in #953
• @duncanmmacleod made their first contribution in #982

Full Changelog: v2.6.0...v2.7.0rc1

v2.7.0rc0

This is a release candidate for Bilby 2.7.0. There are a few major changes/additions in this release along with minor changes and removals.

Major changes

• Likelihood instances can now be called as likelihood.log_likelihood(parameters), see here for more information.
• Support the new API in dynesty=3
• Add a new WaveformGenerator capable of using arbitrary waveform models implemented through the gwsignal waveform interface.

Additions

• Added WeightedCategorical-Prior by @JasperMartins in #893
• ENH: Implement DiscreteValues prior by @unkaktus in #947
• ENH: Allow no parameters as state by @ColmTalbot in #941
• ENH: add support for new dynesty api by @ColmTalbot in #950
• FEAT: add gwsignal waveform generator by @ColmTalbot in #877

Fixes

• BUG: Fix sampling efficiency warning by @fgittins in #953
• TYPO: fix random call in example by @ColmTalbot in #973
• BUG: Fix matched filter SNR calculation in time domain injection by @ColmTalbot in #957
• TST: mark whitened strain tests as flaky by @mj-will in #987
• MAINT: update file extension logic by @mj-will in #962

Changes

• ENH: plot_multiple: Allow plotting onto user-defined figure by @unkaktus in #946
• MAINT: change RNG imports by @mj-will in #943
• MAINT: np.trapz -> np.trapezoid by @ColmTalbot in #974
• MAINT: define Planck15-LAL cosmology using LAL constants by @mj-will in #932
• DEV: make sure all priors return float when needed by @ColmTalbot in #979
• Replace pytables with h5py by @duncanmmacleod in #982

Deprecations

• DEP: deprecate dnest4 interface by @ColmTalbot in #980

Removed

• MAINT: remove unsupported roq json weight file format by @ColmTalbot in #945

Other changes

• DOC: Fixed README link for opening bilby_pipe issues by @mick-wright in #922
• DOC: Correct docs for use_ratio argument of run_sampler by @mattpitkin in #958
• CI: add merge_group to enable merge queue by @mj-will in #983
• TST: remove dnest4 import test by @ColmTalbot in #984
• BLD: migrate to pyproject.toml by @mj-will in #952
• BLD: use release branches by @mj-will in #954

New Contributors

• @fgittins made their first contribution in #953
• @mattpitkin made their first contribution in #958
• @duncanmmacleod made their first contribution in #982

Full Changelog: v2.6.0...v2.7.0rc0

v2.6.0

We noticed a longstanding mis-definition in the GravitationalWaveTransient likelihood that caused the PSD to be
incorrectly normalised when using time domain data with a Tukey window (#869).
By extension, the SNR and log-likelihood is overestimated. While the peak of the likelihood is in
the correct place, the posterior is overly constrained and due to prior effects may not peak in the
correct place.

The specific form of the previous version of the likelihood is

$$ \hat{p}(d | \theta) = \frac{p(d | theta)}{\beta} $$

where $\beta$ is the power loss factor in the Tukey window ($\beta = 1 - 5 \alpha / 8$).

The impact of this bug is most pronounced for short-duration signals, with a long turn on for the Tukey window.
To reproduce the old behaviour, users can set the environment variable BILBY_INCORRECT_PSD_NORMALIZATION=TRUE.

Fixed

• Fix PSD normalisation in GravitationalWaveTransient (#967)

v2.5.2

Release of bilby v2.5.2

This bug fix release resolves an issue introduced in v2.5.1 that caused problems when reading result files.

Fixed

• Fix the filename extension precedence for result files (#960)

Full Changelog: v2.5.1...v2.5.2

v2.5.2rc1

First release candidate for bilby v2.5.2

This is bug fix release.

Fixed

• Fix the filename extension precedence for result files (#960)

v2.5.1

[2.5.1]

This is a bugfix release, most importantly fixing a new bug when reconstructing marginalized parameters for an analysis with an injection.

Changed

• Pin dynesty to version < 2.2 (#949)

Fixed

• Enable printing dlogZ values below 1e-3 with dynesty (#936)
• Fix how injection parameters are handled in parameter conversion to avoid bugs with parameter reconstruction in run_sampler (#931)
• Fix time_reference check in _generate_all_cbc_parameters (#930)
• Ensure output directory exists when performing reweighting with get_weights_for_reweighting (#923)

New Contributors

• @unkaktus made their first contribution in #936
• @Rhiannon-Udall made their first contribution in #930

Full Changelog: v2.5.0...v2.5.1