NumPy-QuadDType v0.2.2

• [ENH] Prioritizes system-wide dependencies, ensuring fallback to meson wrap ones if needed @SwayamInSync

Full Changelog: quaddtype-v0.2.1...quaddtype-v0.2.2

NumPy-QuadDType v0.2.1

• [Fix] multiple copies of the OpenMP runtime initialization issue by @SwayamInSync #169
• [Fix] Don't dereference pointer before knowing it is not NULL by @WarrenWeckesser #167

Full Changelog: quaddtype-v0.2.0...quaddtype-v0.2.1

NumPy-QuadDType v0.2.0

• Improved Packaging
• Cross platform precision improvements1

Contributions

• [Fix] Same Subnormal value over all platform by @SwayamInSync in #141
• Don't suggest -latomic in default build instructions by @ngoldbaum in #142
• Bundle SLEEF and submodules using meson wrap [sdist compatible] by @SwayamInSync in #143
• Implement cast support for ubyte and half by @juntyr in #145
• Make void to quad give error than segfault by @SwayamInSync in #147
• Fixing bad union corruption by @SwayamInSync in #148
• Deprecating macos-13, using macos-14 and 15 by @SwayamInSync in #149
• CI: Windows Pytest fix by @SwayamInSync in #150

Full Changelog: quaddtype-v0.1.0...quaddtype-v0.2.0

NumPy-QuadDType v0.1.0

• Supporting good set of ufuncs
• Extendeding Python support to 3.13t and 3.14t
• Efficient Linear Algebra operations
• More tests

Contributions

• Adding Quaddtype by @SwayamInSync in #98
• Updating Quaddtype by @SwayamInSync in #102
• Add install from source instructions by @mhvk in #104
• fix nan comparison test by @ngoldbaum in #108
• Packaging QuadDType by @SwayamInSync in #100
• NaN tests + Workflow + README fixes by @SwayamInSync in #114
• Test and fix comparisons with NaNs by @juntyr in #115
• Feat: Adding Linear Algebra Dot operation support by @SwayamInSync in #116
• Expose smallest subnormal constant and add tests for consts by @juntyr in #117
• Implement copysign, sign, and signbit ufuncs by @juntyr in #122
• Implement isfinite, isinf, isnan ufuncs by @juntyr in #121
• Implement the fmin and fmax ufuncs by @juntyr in #123
• Implement the reciprocal ufunc by @juntyr in #124
• Fixing mod's behaviour to numpy from IEEE by @SwayamInSync in #125
• Adding tests for uncovered ufuncs by @SwayamInSync in #127
• Add bits, precision, resolution constants by @juntyr in #132
• Resolving padding issues when fractonal part < SLEEF_QUAD_DIG by @SwayamInSync in #133
• Implement hyperbolic ufuncs by @juntyr in #128
• Extending the python support to 3.13 and 3.14 by @SwayamInSync in #134
• Building and uploading sdist along with prebuilt wheels by @SwayamInSync in #135
• Macos-14 runner fix by @SwayamInSync in #138
• Fix rint for near-halfway cases by @juntyr in #137
• MacOS-13 CI fix and Version bump from 0.0.1 to 0.1.0 by @SwayamInSync in #139

New Contributors

• @SwayamInSync made their first contribution in #98
• @mhvk made their first contribution in #104
• @juntyr made their first contribution in #115

Full Changelog: https://github.com/numpy/numpy-user-dtypes/commits/quaddtype-v0.1.0