v2026.5.0

What's Changed

• perf: make test suite snappier and clean up some testing things by @beckermr in #171
• feat: add to_galsim methods for positions and bounds by @beckermr in #172
• prod: lower bound for jax to 0.6 by @beckermr in #173
• test: retry once for some flaky tests by @beckermr in #174
• feat: add walltime benchmark in addition to simulation by @beckermr in #175
• fix: link to CODE_OF_CONDUCT.md in README is busted by @beckermr in #177
• feat: add to_galsim method for Shear, Angle, CelestialCoord, and WCS classes by @beckermr in #176
• fix: use adaptive precision for moffat computation by @beckermr in #181
• add to_galsim for Image class, demo1.py, & demo2.py by @b-remy in #163
• feat: add benchmark for moffat as PSF by @beckermr in #186
• feat: add benchmark for gradient w/ moffat by @beckermr in #187
• perf: use an interpolant for maxk for Moffat profile, drop support for trunc option by @beckermr in #196
• fix: missed notebook contents! by @beckermr in #197
• fix: nan in derivs wrt params for Moffat by @beckermr in #194
• perf: use interpolant and asymptotic expansion for Moffat by @beckermr in #198
• feat: add stop gradient calls for certain cases for beta/nu in Moffat and Spergel plus docs by @beckermr in #199
• fix: do not use lazy properties or workspaces by @beckermr in #200
• test: add vmap gradient test for Gaussian object by @beckermr in #202
• feat: complete to-from-galsim interface for gsparams by @beckermr in #206
• chore(deps): bump the github-actions group with 4 updates by @dependabot[bot] in #207
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #208
• test: move testing submodule to fork by @beckermr in #213
• fix: ensure that we properly unpack a tuple in string formatting by @beckermr in #214
• add requires-python field in pyproject toml by @ismael-mendoza in #215
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #217
• fix: use dynamic slice for set and get ops for images by @beckermr in #212
• test: update galsim submodule for new interpolated image tests by @beckermr in #219
• test: update tests with latest commits from main by @beckermr in #220
• fix: make stepk match galsim by @beckermr in #221
• test: no need to set stepk anymore by @beckermr in #223
• feat: investigate maxk differences and add galsim algorithm for maxk by @beckermr in #224
• Sphinx documentation by @ismael-mendoza in #226
• fix: update JAX-GalSim Sharp Bits section title by @beckermr in #227
• feat: update API list in docs with script by @beckermr in #228
• fix: use slower maxk comp to match galsim by @beckermr in #229

Full Changelog: v2026.2.0...v2026.5.0

v2026.2.0

What's Changed

• Adding automated test suite mechanism to branch #5 by @EiffL in #14
• Adding minimal GSObject, Sum, Gaussian, Exponentional by @EiffL in #7
• Adding CONTRIBUTING document and related infrastructure by @EiffL in #8
• Adds jittable Position object by @EiffL in #11
• Add black test to CI by @ismael-mendoza in #34
• Adds jittable wcs and Transformation by @EiffL in #12
• Adds jittable Bounds by @EiffL in #15
• Adds jittable Image by @EiffL in #16
• WIP: Implementing drawing functions, and testing jit patterns by @EiffL in #21
• Implementing shear by @ismael-mendoza in #41
• Pixel by @ismael-mendoza in #43
• Restoring canonical GalSim Image API by @EiffL in #47
• Implementation of k-space convolution by @EiffL in #44
• STY clean out a few flake8 warnings by @beckermr in #55
• PROD clean out pre-commit config and add ruff settings by @beckermr in #57
• BUG make sure pytest can run -vvs by @beckermr in #58
• ENH make sure eval(repr(o)) == o and pickling works by @beckermr in #59
• ENH add fits readers by @beckermr in #61
• BUG fix accuracy loss in sheared image test by @beckermr in #62
• BUG make sure to add to image in-place by @beckermr in #63
• Moffat by @jecampagne in #49
• isort by @ismael-mendoza in #67
• ENH API tests and fixes by @beckermr in #68
• ENH add Angle and CelestialCoord by @beckermr in #69
• ENH add deconv by @beckermr in #70
• ENH add random deviates by @beckermr in #72
• TST fix CI tests after rebase on main by @beckermr in #73
• ENH add Interpolant classes by @beckermr in #56
• ENH add InterpolatedImage by @beckermr in #60
• ENH enable noise fields and fix drawing bugz by @beckermr in #79
• ENH add GSFitsWCS by @beckermr in #80
• ENH add photon shooting by @beckermr in #82
• Spergel by @jecampagne in #86
• ENH update to latest JAX version and slight refactor in test suite by @ismael-mendoza in #94
• REF use our own implements decorator by @beckermr in #95
• PERF block reduce in wrapping by @beckermr in #97
• test: add benchmarks for interpolated images by @beckermr in #100
• feat: add more benchmarks and clean up code by @beckermr in #102
• fix: ensure jax caches are always cleared for compile benchmarks by @beckermr in #103
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #105
• perf: use vmap for iimage by @beckermr in #104
• refactor: do interpolation class init on the host by @beckermr in #106
• perf: faster xval for Lanczos interp by @beckermr in #107
• perf: faster kval for Lanczos interpolated images via cached splines by @beckermr in #101
• refactor: move to ruff + ruff-format by @beckermr in #109
• feat: update for coord 1.4 and point tests at main branch by @beckermr in #110
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #112
• chore: remove old pylint config by @beckermr in #115
• chore: move to pyproject.toml by @beckermr in #113
• doc: start a docs update by @beckermr in #114
• fix: consistent doc string wrapping by @beckermr in #116
• chore: code clean up by @beckermr in #117
• feat: add int1d by @beckermr in #118
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #119
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #120
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #121
• fix: make sure bounds for images are set consistently when using center with nx/ny by @beckermr in #122
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #126
• feat: add benchmarks for spergel profile by @beckermr in #125
• Bump the github-actions group with 2 updates by @dependabot[bot] in #127
• feat: add benchmarks for class init by @beckermr in #128
• perf: use pseudo-Pade approximation for the spergel HLR computation by @beckermr in #130
• feat: add benchmark for RNG discard operation by @beckermr in #132
• feat: add benchmark and test of flux frac for interp images by @beckermr in #133
• perf: optimize flux frac for interpolated images by removing fori_loop by @beckermr in #134
• perf: remove fori_loops for polynomial eval by @beckermr in #135
• test: add benchmark for fits by @beckermr in #137
• perf: remove fori loop for fits wcs by @beckermr in #136
• test: add benchmark for moffat init by @beckermr in #139
• perf: try a purely compiled moffat init by @beckermr in #140
• test: make test suite faster with pytest-split and add jit compile optimization for fits wcs by @beckermr in #138
• perf: try using fori_loop but unrolling everything by @beckermr in #141
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #142
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #143
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #144
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #145
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #146
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #147
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #148
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #150
• chore(deps): bump actions/checkout from 4 to 5 in the github-actions group by @dependabot[bot] in #151
• chore(deps): bump the github-actions group with 2 updates by @dependabot[bot] in #152
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #153
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #154
• chore(deps): bump actions/checkout from 5 to 6 in the github-actions group by @dependabot[bot] in #155
• [pre...

Version used for Shear Inference OJA paper

This tag represents the version of jax galsim used to obtain results for the paper "Differentiable Forward Modeling for Efficient and Accurate Shear Inference".