|
| 1 | +.. currentmodule:: numpy |
| 2 | + |
| 3 | +========================== |
| 4 | +NumPy 1.26.3 Release Notes |
| 5 | +========================== |
| 6 | + |
| 7 | +NumPy 1.26.3 is a maintenance release that fixes bugs and regressions |
| 8 | +discovered after the 1.26.2 release. The most notable changes are the f2py bug |
| 9 | +fixes. The Python versions supported by this release are 3.9-3.12. |
| 10 | + |
| 11 | + |
| 12 | +Compatibility |
| 13 | +============= |
| 14 | + |
| 15 | +``f2py`` will no longer accept ambiguous ``-m`` and ``.pyf`` CLI combinations. |
| 16 | +When more than one ``.pyf`` file is passed, an error is raised. When both ``-m`` |
| 17 | +and a ``.pyf`` is passed, a warning is emitted and the ``-m`` provided name is |
| 18 | +ignored. |
| 19 | + |
| 20 | + |
| 21 | +Improvements |
| 22 | +============ |
| 23 | + |
| 24 | +``f2py`` now handles ``common`` blocks which have ``kind`` specifications from |
| 25 | +modules. This further expands the usability of intrinsics like |
| 26 | +``iso_fortran_env`` and ``iso_c_binding``. |
| 27 | + |
| 28 | + |
| 29 | +Contributors |
| 30 | +============ |
| 31 | + |
| 32 | +A total of 18 people contributed to this release. People with a "+" by their |
| 33 | +names contributed a patch for the first time. |
| 34 | + |
| 35 | +* @DWesl |
| 36 | +* @Illviljan |
| 37 | +* Alexander Grund |
| 38 | +* Andrea Bianchi + |
| 39 | +* Charles Harris |
| 40 | +* Daniel Vanzo |
| 41 | +* Johann Rohwer + |
| 42 | +* Matti Picus |
| 43 | +* Nathan Goldbaum |
| 44 | +* Peter Hawkins |
| 45 | +* Raghuveer Devulapalli |
| 46 | +* Ralf Gommers |
| 47 | +* Rohit Goswami |
| 48 | +* Sayed Adel |
| 49 | +* Sebastian Berg |
| 50 | +* Stefano Rivera + |
| 51 | +* Thomas A Caswell |
| 52 | +* matoro |
| 53 | + |
| 54 | + |
| 55 | +Pull requests merged |
| 56 | +==================== |
| 57 | +A total of 42 pull requests were merged for this release. |
| 58 | + |
| 59 | +* `#25130 <https://github.com/numpy/numpy/pull/25130>`__: MAINT: prepare 1.26.x for further development |
| 60 | +* `#25188 <https://github.com/numpy/numpy/pull/25188>`__: TYP: add None to ``__getitem__`` in ``numpy.array_api`` |
| 61 | +* `#25189 <https://github.com/numpy/numpy/pull/25189>`__: BLD,BUG: quadmath required where available [f2py] |
| 62 | +* `#25190 <https://github.com/numpy/numpy/pull/25190>`__: BUG: alpha doesn't use REAL(10) |
| 63 | +* `#25191 <https://github.com/numpy/numpy/pull/25191>`__: BUG: Fix FP overflow error in division when the divisor is scalar |
| 64 | +* `#25192 <https://github.com/numpy/numpy/pull/25192>`__: MAINT: Pin scipy-openblas version. |
| 65 | +* `#25201 <https://github.com/numpy/numpy/pull/25201>`__: BUG: Fix f2py to enable use of string optional inout argument |
| 66 | +* `#25202 <https://github.com/numpy/numpy/pull/25202>`__: BUG: Fix -fsanitize=alignment issue in numpy/_core/src/multiarray/arraytypes.c.src |
| 67 | +* `#25203 <https://github.com/numpy/numpy/pull/25203>`__: TST: Explicitly pass NumPy path to cython during tests (also... |
| 68 | +* `#25204 <https://github.com/numpy/numpy/pull/25204>`__: BUG: fix issues with ``newaxis`` and ``linalg.solve`` in ``numpy.array_api`` |
| 69 | +* `#25205 <https://github.com/numpy/numpy/pull/25205>`__: BUG: Disallow shadowed modulenames |
| 70 | +* `#25217 <https://github.com/numpy/numpy/pull/25217>`__: BUG: Handle common blocks with kind specifications from modules |
| 71 | +* `#25218 <https://github.com/numpy/numpy/pull/25218>`__: BUG: Fix moving compiled executable to root with f2py -c on Windows |
| 72 | +* `#25219 <https://github.com/numpy/numpy/pull/25219>`__: BUG: Fix single to half-precision conversion on PPC64/VSX3 |
| 73 | +* `#25227 <https://github.com/numpy/numpy/pull/25227>`__: TST: f2py: fix issue in test skip condition |
| 74 | +* `#25240 <https://github.com/numpy/numpy/pull/25240>`__: Revert "MAINT: Pin scipy-openblas version." |
| 75 | +* `#25249 <https://github.com/numpy/numpy/pull/25249>`__: MAINT: do not use ``long`` type |
| 76 | +* `#25377 <https://github.com/numpy/numpy/pull/25377>`__: TST: PyPy needs another gc.collect on latest versions |
| 77 | +* `#25378 <https://github.com/numpy/numpy/pull/25378>`__: CI: Install Lapack runtime on Cygwin. |
| 78 | +* `#25379 <https://github.com/numpy/numpy/pull/25379>`__: MAINT: Bump conda-incubator/setup-miniconda from 2.2.0 to 3.0.1 |
| 79 | +* `#25380 <https://github.com/numpy/numpy/pull/25380>`__: BLD: update vendored Meson for AIX shared library fix |
| 80 | +* `#25419 <https://github.com/numpy/numpy/pull/25419>`__: MAINT: Init ``base`` in cpu_avx512_kn |
| 81 | +* `#25420 <https://github.com/numpy/numpy/pull/25420>`__: BUG: Fix failing test_features on SapphireRapids |
| 82 | +* `#25422 <https://github.com/numpy/numpy/pull/25422>`__: BUG: Fix non-contiguous memory load when ARM/Neon is enabled |
| 83 | +* `#25428 <https://github.com/numpy/numpy/pull/25428>`__: MAINT,BUG: Never import distutils above 3.12 [f2py] |
| 84 | +* `#25452 <https://github.com/numpy/numpy/pull/25452>`__: MAINT: make the import-time check for old Accelerate more specific |
| 85 | +* `#25458 <https://github.com/numpy/numpy/pull/25458>`__: BUG: fix macOS version checks for Accelerate support |
| 86 | +* `#25465 <https://github.com/numpy/numpy/pull/25465>`__: MAINT: Bump actions/setup-node and larsoner/circleci-artifacts-redirector-action |
| 87 | +* `#25466 <https://github.com/numpy/numpy/pull/25466>`__: BUG: avoid seg fault from OOB access in RandomState.set_state() |
| 88 | +* `#25467 <https://github.com/numpy/numpy/pull/25467>`__: BUG: Fix two errors related to not checking for failed allocations |
| 89 | +* `#25468 <https://github.com/numpy/numpy/pull/25468>`__: BUG: Fix regression with ``f2py`` wrappers when modules and subroutines... |
| 90 | +* `#25475 <https://github.com/numpy/numpy/pull/25475>`__: BUG: Fix build issues on SPR |
| 91 | +* `#25478 <https://github.com/numpy/numpy/pull/25478>`__: BLD: fix uninitialized variable warnings from simd/neon/memory.h |
| 92 | +* `#25480 <https://github.com/numpy/numpy/pull/25480>`__: BUG: Handle ``iso_c_type`` mappings more consistently |
| 93 | +* `#25481 <https://github.com/numpy/numpy/pull/25481>`__: BUG: Fix module name bug in signature files [urgent] [f2py] |
| 94 | +* `#25482 <https://github.com/numpy/numpy/pull/25482>`__: BUG: Handle .pyf.src and fix SciPy [urgent] |
| 95 | +* `#25483 <https://github.com/numpy/numpy/pull/25483>`__: DOC: ``f2py`` rewrite with ``meson`` details |
| 96 | +* `#25485 <https://github.com/numpy/numpy/pull/25485>`__: BUG: Add external library handling for meson [f2py] |
| 97 | +* `#25486 <https://github.com/numpy/numpy/pull/25486>`__: MAINT: Run f2py's meson backend with the same python that ran... |
| 98 | +* `#25489 <https://github.com/numpy/numpy/pull/25489>`__: MAINT: Update ``numpy/f2py/_backends`` from main. |
| 99 | +* `#25490 <https://github.com/numpy/numpy/pull/25490>`__: MAINT: Easy updates of ``f2py/*.py`` from main. |
| 100 | +* `#25491 <https://github.com/numpy/numpy/pull/25491>`__: MAINT: Update crackfortran.py and f2py2e.py from main |
| 101 | + |
0 commit comments