v3.0.2

What's Changed

• Implementation of Cov NG terms by @carlosggarcia in #869
• Integration of N5K winners into CCL by @slosar in #1092
• Bacco Lagrangian bias expansion by @matteozennaro in #1139
• Background Calculations by @hsinfan1996 in #1149
• Documentation fixes by @damonge in #1122
• Fix GSL integration abort by @tilmantroester in #1155
• Add to_dict utility method by @tilmantroester in #1160

Full Changelog: v3.0.0...v3.0.2

v3.0.1

What's Changed

• Implementation of Cov NG terms by @carlosggarcia in #869
• Integration of N5K winners into CCL by @slosar in #1092
• Bacco Lagrangian bias expansion by @matteozennaro in #1139
• Background Calculations by @hsinfan1996 in #1149
• Documentation fixes by @damonge in #1122
• Fix GSL integration abort by @tilmantroester in #1155
• Add to_dict utility method by @tilmantroester in #1160

Full Changelog: v3.0.0...v3.0.1

First v3 version

What's Changed

• API changes in v3 mode (#1087)
• Power spectrum emulators (#1094, #1102, #1103).
• Improvements to matter density profiles by @hsinfan1996 (#1096, #1093)
• LPT NL PT by @anicola in #1097
• Better treatment of halo model overdensity parameters (#1101, #1100)
• More and better treatment of baryonic effect models (#1108, #1110, #1113)
• Allow sigma8 as an input when using CAMB for nonlinear power spectra by @tilmantroester in #1106
• Bocquet 2020 mass function emulator by @damonge in #1115
• Container for MG params by @tilmantroester in #1119
• Improved documentation (#1116)

New Contributors

• @matteozennaro made their first contribution in #1094
• @anicola made their first contribution in #1097

Full Changelog: v2.8.0...v3.0.0

Last v2-compatible version, with v3 API

This is the last version of CCL that will be compatible with the v2 API. This version is also able to function under much of the basic v3 API, and will in fact throw warnings when used with the v2 API. This is intended as a way to soften the transition from v2 to v3. Users who want to benefit from it can migrate to this version first, fix their code until there are no more warnings, and then they'll be ready to use v3 proper. If you would rather stick to v2 for a while longer, without worrying about this, consider using the last pure v2 version (v2.7.0), which contains none of these warnings.

What's Changed

• Refactor pyccl under an abstract CCLObject. by @nikfilippas in #934
• Bump Python version to 3.8 by @nikfilippas in #1031
• RTD fix by @damonge in #1032
• Optimize HM autocorrelation power spectra by @nikfilippas in #891
• Fix incorrect Python version in CI by @nikfilippas in #1037
• Cleanup of conda-incubator & replace deprecated pkg_resources by @nikfilippas in #1045
• Fix error message for incorrect calls with unlock_instance by @nikfilippas in #1044
• Unfreeze option for physical_constants by @nikfilippas in #1050
• Implement separate __eq__ methods by @nikfilippas in #1033
• v2.final by @nikfilippas in #1038
• Docs fix by @damonge in #1081
• API restoration by @damonge in #1083
• IA profile v3 by @damonge in #1074
• Lightweight docs revamp for v2.last by @damonge in #1082
• Bug from sed by @damonge in #1086
• Changelog update by @damonge in #1089

Full Changelog: v2.7.0...v2.8.0

Last pure v2 version

This is the last version of CCL using the pure v2 API, before introducing any of the new underlying structure and deprecation messages in the buildup to v3. One last v2 release will be made after this one, in which users will be able to use either the v2 or v3 APIs, with deprecation warnings thrown if using v2.

v2.6.1

What's Changed

• Numpy 1.24 compatibility by @hsinfan1996 in #1012
• Fix OpenMP again by @tilmantroester in #1017

Full Changelog: v2.6...v2.6.1

v2.6

What's Changed

• Fixed new flake8 errors by @damonge in #967
• Add test to profiles on squeezed shapes and fix Hernquist _fourier_analytic by @hsinfan1996 in #966
• Enhancements for Pk2D, Tk3D, Tracer. by @nikfilippas in #923
• Useful error for transfer functions with non-monotonic scale factors by @damonge in #971
• Better docs for correlations by @damonge in #974
• Stop supporting Python 3.6 by @nikfilippas in #975
• HaloProfileGNFW bugfix for c500 update. by @nikfilippas in #969
• Compute non-linear power spectrum for (trf, mps) == (not camb, camb) combinations. by @nikfilippas in #979
• GitHub Actions: Fix class installation by @hsinfan1996 in #984
• Add option to make einasto alpha a free parameter by @hsinfan1996 in #987
• Fix free Einasto alpha dimension by @hsinfan1996 in #989
• MG parameters in CosmologyCalculator by @nikfilippas in #977
• MG test fix by @damonge in #1006
• Replace m2r2 with sphinx_mdinclude by @hsinfan1996 in #1005
• Add functions to check if OpenMP is working by @hsinfan1996 in #1000
• Enhancement to the perturbative modelling by @tilmantroester in #1002

Full Changelog: v2.5.1...v2.6

v2.5.1

What's Changed

• Analytic Einasto and Hernquist normalization by @hsinfan1996 in #956
• General bugfixes & enhancements. by @nikfilippas in #925
• Angular distance for one lens and multiple sources by @damonge in #950
• Bug in spline checks by @damonge in #961
• Halofit check for range of background wa != 0. by @tilmantroester in #957
• Bug in magnification bias by @damonge in #963
• Add analytic _projected, _cumul2d and _fourier to Hernquist profile by @hsinfan1996 in #958
• Tunable CCL global settings at the Python-level by @nikfilippas in #918

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

v2.5.0

What's Changed

• Fixing missing terms in Super Sample Covariance by @carlosggarcia in #941
• Fix status message string initialisation by @tilmantroester in #935
• Fix WL functions in pyccl.halos.profiles by @hsinfan1996 in #943
• 10x speed increase in CIB profile by @nikfilippas in #917
• Super Sample Covariance with linear galaxy bias approximation by @carlosggarcia in #947
• Add action to publish to pypi by @joezuntz in #948

New Contributors

• @carlosggarcia made their first contribution in #941
• @hsinfan1996 made their first contribution in #943

Full Changelog: v2.4.0...v2.5.0

v2.4.0

What's Changed

• Additional PT biases by @damonge in #901
• ISW by @damonge in #861
• CIB profiles by @damonge in #905
• Implement mass-concentrations from Ishiyama et al. (c500) by @nikfilippas in #895
• Update docs and tests to use mead2020_feedback. by @tilmantroester in #921
• Make tracer spline errors more verbose by @tilmantroester in #926
• TST add test for high-z halofit by @beckermr in #932
• c(M) Ishiyama21 speedup by @nikfilippas in #930
• Option to compile CCL with debugging symbols by @tilmantroester in #931
• Allow access to Tk3D spline data by @nikfilippas in #924
• Make tracer creation more robust by @tilmantroester in #928
• Update changelog by @damonge in #939

Full Changelog: v2.3.0...v2.4.0