Releases: igmhub/picca
Louis the Pious
This version primarily includes the faster healpix readin, improvement for the metal matrix as well as Pk1d/cross exposure, and adds IVAR to the delta files even for non-Pk1d analyses.
What's Changed
- Moving constants to using scipy.constants by @iprafols in #1055
- fix raw_io for default delta_lambda by @cramirezpe in #1057
- read version directly from version file by @iprafols in #1058
- Fixed documentation related to projection, and removed innecessary option by @andreufont in #1061
- [bump minor] Cross exposure by @corentinravoux in #1056
- fixing a tile bug for p1d by @corentinravoux in #1066
- Update numpy2.0 [bump minor] by @iprafols in #1070
- [bump minor] Pk1d covariance new by @corentinravoux in #1067
- Fixed error when package downloaded from tarball instead of cloned by @iprafols in #1069
- Fix testing by @Waelthus in #1075
- Average mean pk by @corentinravoux in #1072
- Improvement to the fast metal matrix computation by @julienguy in #1073
- Fast desihealpix [bump minor] by @iprafols in #1071
- [bump minor] add IVAR to all forest outputs by @Waelthus in #1074
Full Changelog: v9.0.0...v9.5.0
Contributors
Assets 2
Charlemagne
Starting with this version and following the DESI Hawaii meeting unblinding, the usage of Y1 unblinded DESI data is allowed. So this is the official version to do unblinded DESI analyses with Y1 data and the starting point for Y3 blinded analyses.
Additionally there are a couple minor bugfixes.
What's Changed
- if no variabes are floated, do not perform fit by @iprafols in #1043
- fix and update BAL wavelengths for Lyb region by @iprafols in #1048
- fixed the coveralls workflow by @iprafols in #1051
- Fix raw mocks bug by @andreicuceu in #1049
- Add flag to turn off non-blind cosmology messages by @andreicuceu in #1050
- [bump major] Unblinding y1 by @andreicuceu in #1054
Full Changelog: v8.3.0...v9.0.0
Contributors
Assets 2
Pepin the Short
This release fixes the metal matrices by a) allowing to only calculate the most relevant species, b) changing how redshifts are reported, i.e. such that the actual absorber redshift is reported
What's Changed
- add relevant metals option to fast_metal_dmat by @calumgordon in #1042
- Redshifts in metal matrices [bump minor] by @julienguy in #1045
Full Changelog: v8.2.0...v8.3.0
Contributors
Assets 2
Childeric III
This version includes the fast computation of metal matrices as well as some bug-fixes and documentation updates.
All baseline v1 analyses of the 2023 KP6 workshop should be produced with this release.
What's Changed
- changes to picca_cf1d by @Waelthus in #1036
- P1d nan fix by @corentinravoux in #1038
- added fast-metals option to picca_metal_dmat [minor] by @calumgordon in #1039
- Update picca_delta_extraction_configuration_tutorial.ipynb by @p-slash in #1041
- Fast computation of metal dmat using stack of delta [bump minor] by @julienguy in #1040
Full Changelog: v8.1.0...v8.2.0
Contributors
Assets 2
Theuderic IV
New minor version mostly because from here on the version will be saved into your .config files.
Also a fix to P1d, this will be used for the final EDR analyses by @corentinravoux.
What's Changed
- Add version to config by @cramirezpe in #1035
- [bump minor] P1d covariance fix by @corentinravoux in #1034
Full Changelog: v8.0.0...v8.1.0
Contributors
Assets 2
Chilperic II
The main purpose of this PR is to finally retire old picca_deltas and fitter2 and io code required for them. These routines have been considered deprecated for > 1yr now (since v5.0.0) and their removal speeds up the automated testing and simplifies moving to new versions, which sometimes led to issues with the old code. If you need to use the old routines for any reason, please downgrade to earlier versions.
What's Changed
- new option per_r_par for covariance smoothing by @julienguy in #1030
- Remove old picca routines [bump major] by @Waelthus in #1033
Full Changelog: v7.2.0...v8.0.0
Contributors
Assets 2
2023 Durham meeting
This version has been used in analyses by @cramirezpe for analyses shown at the 2023 Durham meeting.
Contributors
Assets 2
Dagobert III
The main change for this version is an update of interpolation functions from nearest-neighbor to cubic interpolators. Elsewise mostly small updates and an update of tests to include python 3.11.
What's Changed
- Fix desi tile by @iprafols in #1023
- Fix xcf1d by @iprafols in #1024
- simplified code to compute var stats fits by @iprafols in #1025
- Change delta extraction interpolations to cubic [bump minor] by @andreicuceu in #1008
- Update the github action for tests by @Waelthus in #1028
Full Changelog: v7.1.0...v7.2.0
Contributors
Assets 2
Childebert III
This release is mostly fixing bugs in the DLA model and unifying the relevant DLA constants between picca and qsonic.
Also a bug in reading SDSS data has been fixed and the transition wavelengths for several metal species have been added to the constasts files.
Finally, from now on we automatically bump the patch-level whenever a pull request is merged.
What's Changed
- Add workflow to auto-submit to pypi by @Waelthus in #1006
- Add github action to automatically bump version on PR by @Waelthus in #1011
- Update utils.py by @Andrea-MG in #1012
- adding merge-groups to actions by @Waelthus in #1014
- Don't run CI on pure markdown changes by @Waelthus in #1015
- [no bump] Let's try the merge_groups again... by @Waelthus in #1016
- Update constants.py by @Andrea-MG in #1013
- change_auto_versioning by @Waelthus in #1017
- Correction of CIII in constants.py by @Andrea-MG in #1020
- SDSS fix by @armengau in #1018
- Update bump_version.yml by @Waelthus in #1021
- Fix dla model by @cramirezpe in #1022
- [bump minor] add tags for major/minor versions by @Waelthus in #1026
New Contributors
- @Andrea-MG made their first contribution in #1012
Full Changelog: v7.0.0...v7.1.0
Contributors
Assets 2
Clovis IV
Updates in several parts of the code on the way to Y1, see below.
There are some changes regarding detailed definitions of the wavelength grids to improve performance. Therefore, results will change very slightly compared to previous versions. This has no measureable effect on the resulting correlations.
For Pk1d backwards incompatible changes have been done regarding the resolution correction based on the resolution matrix, the approach in here does the correct thing for data, pixel-level sims and new quickquasar mocks (previous versions were only good for quickquasars because of a bug in there that has been fixed).
What's Changed
- Add new Planck18 template for DESI by @andreicuceu in #966
- Fix deltas rebin by @cramirezpe in #967
- Pk1d zchunks by @armengau in #968
- Couple of very minor fixes by @iprafols in #969
- fix so that we can run on eBOSS mocks from quickquasrs by @iprafols in #965
- True cont update by @alxogm in #972
- Change LASTNIGHT for COADD_LASTNIGHT by @alxogm in #973
- Change default redshift evolution to match QSO by @andreicuceu in #974
- Update picca_Pk1D.py by @alxogm in #976
- add cuts in quasar redshift by @cramirezpe in #977
- Fix data model by @cramirezpe in #979
- Fix blinding for table format by @cramirezpe in #980
- Read lowercase calib fits by @cramirezpe in #982
- fix read redshift through z_key by @cramirezpe in #981
- fix_issue985 by @iprafols in #986
- update reading var stats from fits file by @cramirezpe in #987
- Fix the resolution correction based on pixel level simulation findings by @corentinravoux in #988
- Pk1d covariance by @corentinravoux in #975
- Pk1d test by @armengau in #991
- added tests for DLA and BAL masking by @iprafols in #989
- Cosmo class fix by @iprafols in #996
- modified weights by @iprafols in #998
- astronomical object uses rad as units (not deg) by @cramirezpe in #983
- Fix raw io by @andreicuceu in #999
- Optimize masks by @iprafols in #984
- fix deprecated numpy/numba usages by @Waelthus in #1002
- Optimize find bins fix issue 993 by @iprafols in #1000
- skip apply_masks and apply_corrections when not needed by @p-slash in #1004
- Fix p1d weights by @armengau in #1005
- Pk1d diff noise estimators by @corentinravoux in #997
- Pk1d covariance improvements by @corentinravoux in #1003
New Contributors
Full Changelog: v6.1...v7.0.0
What's Changed
- Add new Planck18 template for DESI by @andreicuceu in #966
- Fix deltas rebin by @cramirezpe in #967
- Pk1d zchunks by @armengau in #968
- Couple of very minor fixes by @iprafols in #969
- fix so that we can run on eBOSS mocks from quickquasrs by @iprafols in #965
- True cont update by @alxogm in #972
- Change LASTNIGHT for COADD_LASTNIGHT by @alxogm in #973
- Change default redshift evolution to match QSO by @andreicuceu in #974
- Update picca_Pk1D.py by @alxogm in #976
- add cuts in quasar redshift by @cramirezpe in #977
- Fix data model by @cramirezpe in #979
- Fix blinding for table format by @cramirezpe in #980
- Read lowercase calib fits by @cramirezpe in #982
- fix read redshift through z_key by @cramirezpe in #981
- fix_issue985 by @iprafols in #986
- update reading var stats from fits file by @cramirezpe in #987
- Fix the resolution correction based on pixel level simulation findings by @corentinravoux in #988
- Pk1d covariance by @corentinravoux in #975
- Pk1d test by @armengau in #991
- added tests for DLA and BAL masking by @iprafols in #989
- Cosmo class fix by @iprafols in #996
- modified weights by @iprafols in #998
- astronomical object uses rad as units (not deg) by @cramirezpe in #983
- Fix raw io by @andreicuceu in #999
- Optimize masks by @iprafols in #984
- fix deprecated numpy/numba usages by @Waelthus in #1002
- Optimize find bins fix issue 993 by @iprafols in #1000
- skip apply_masks and apply_corrections when not needed by @p-slash in #1004
- Fix p1d weights by @armengau in #1005
- Pk1d diff noise estimators by @corentinravoux in #997
- Pk1d covariance improvements by @corentinravoux in #1003
New Contributors
Full Changelog: v6.1...v7.0.0
