diff options
| -rw-r--r-- | .mailmap | 6 | ||||
| -rw-r--r-- | doc/changelog/1.23.0-changelog.rst | 660 | ||||
| -rw-r--r-- | doc/source/release/1.23.0-notes.rst | 387 |
3 files changed, 1041 insertions, 12 deletions
@@ -120,7 +120,8 @@ Brett R Murphy <bmurphy@enthought.com> Brian Soto <iambriansoto@gmail.com> Brian Soto <iambriansoto@gmail.com> <Iamsoto@users.noreply.github.com> Brian Soto <iambriansoto@gmail.com> <theintrocode@gmail.com> -Brigitta Sipocz <bsipocz@gmail.com> <b.sipocz@gmail.com> +Brigitta Sipőcz <bsipocz@gmail.com> +Brigitta Sipőcz <bsipocz@gmail.com> <b.sipocz@gmail.com> Bryan Van de Ven <bryanv@continuum.io> Bryan Van de Ven <bryan@Laptop-3.local> Bryan Van de Ven <bryanv@continuum.io> Bryan Van de Ven <bryan@laptop.local> Brénainn Woodsend <bwoodsend@gmail.com> @@ -267,6 +268,8 @@ Jérome Eertmans <jeertmans@icloud.com> Jérôme Richard <jeromerichard111@msn.com> <ubuntu@ip-172-31-17-195.eu-west-3.compute.internal> Jerome Kelleher <jerome.kelleher@ed.ac.uk> Jessi J Zhao <35235453+jessijzhao@users.noreply.github.com> +Jhong-Ken Chen (陳仲肯) <kenny.kuo.fs@gmail.com> +Jhong-Ken Chen (陳仲肯) <kenny.kuo.fs@gmail.com> <37182101+kennychenfs@users.noreply.github.com> Johannes Hampp <johannes.hampp@zeu.uni-giessen.de> <42553970+euronion@users.noreply.github.com> Johannes Schönberger <hannesschoenberger@gmail.com> <jschoenberger@demuc.de> Johann Faouzi <johann.faouzi@gmail.com> <johann.faouzi@icm-institute.org> @@ -528,4 +531,3 @@ Zieji Pohz <poh.ziji@gmail.com> Zieji Pohz <poh.ziji@gmail.com> <8103276+zjpoh@users.noreply.github.com> Zolboo Erdenebaatar <erdenebz@dickinson.edu> Zolisa Bleki <zolisa.bleki@gmail.com> <44142765+zoj613@users.noreply.github.com> -陳仲肯 (Chen, Jhong-Ken) <kenny.kuo.fs@gmail.com> diff --git a/doc/changelog/1.23.0-changelog.rst b/doc/changelog/1.23.0-changelog.rst new file mode 100644 index 000000000..3ff0dcfd5 --- /dev/null +++ b/doc/changelog/1.23.0-changelog.rst @@ -0,0 +1,660 @@ + +Contributors +============ + +A total of 151 people contributed to this release. People with a "+" by their +names contributed a patch for the first time. + +* @DWesl +* @GalaxySnail + +* @code-review-doctor + +* @h-vetinari +* Aaron Meurer +* Alexander Shadchin +* Alexandre de Siqueira +* Allan Haldane +* Amrit Krishnan +* Andrei Batomunkuev +* Andrew J. Hesford + +* Andrew Murray + +* Andrey Andreyevich Bienkowski + +* André Elimelek de Weber + +* Andy Wharton + +* Arryan Singh +* Arushi Sharma +* Bas van Beek +* Bharat Raghunathan +* Bhavuk Kalra + +* Brigitta Sipőcz +* Brénainn Woodsend + +* Burlen Loring + +* Caio Agiani + +* Charles Harris +* Chiara Marmo +* Cornelius Roemer + +* Dahyun Kim + +* Damien Caliste +* David Prosin + +* Denis Laxalde +* Developer-Ecosystem-Engineering +* Devin Shanahan + +* Diego Wang + +* Dimitri Papadopoulos Orfanos +* Ding Liu + +* Diwakar Gupta + +* Don Kirkby + +* Emma Simon + +* Eric Wieser +* Evan Miller + +* Evgeni Burovski +* Evgeny Posenitskiy + +* Ewout ter Hoeven + +* Felix Divo +* Francesco Andreuzzi + +* Ganesh Kathiresan +* Gaëtan de Menten +* Geoffrey Gunter + +* Hans Meine +* Harsh Mishra + +* Henry Schreiner +* Hood Chatham + +* I-Shen Leong +* Ilhan Polat +* Inessa Pawson +* Isuru Fernando +* Ivan Gonzalez + +* Ivan Meleshko + +* Ivan Yashchuk + +* Janus Heide + +* Jarrod Millman +* Jason Thai + +* Jeremy Volkman + +* Jesús Carrete Montaña + +* Jhong-Ken Chen (陳仲肯) + +* John Kirkham +* John-Mark Gurney + +* Jonathan Deng + +* Joseph Fox-Rabinovitz +* Jouke Witteveen + +* Junyan Ou + +* Jérôme Richard + +* Kassian Sun + +* Kazuki Sakamoto + +* Kenichi Maehashi +* Kevin Sheppard +* Kilian Lieret + +* Kushal Beniwal + +* Leo Singer +* Logan Thomas + +* Lorenzo Mammana + +* Margret Pax +* Mariusz Felisiak + +* Markus Mohrhard + +* Mars Lee +* Marten van Kerkwijk +* Masamichi Hosoda + +* Matthew Barber +* Matthew Brett +* Matthias Bussonnier +* Matthieu Darbois +* Matti Picus +* Melissa Weber Mendonça +* Michael Burkhart + +* Morteza Mirzai + +* Motahhar Mokf + +* Muataz Attaia + +* Muhammad Motawe + +* Mukulika Pahari +* Márton Gunyhó + +* Namami Shanker + +* Nihaal Sangha + +* Niyas Sait +* Omid Rajaei + +* Oscar Gustafsson + +* Ovee Jawdekar + +* P. L. Lim + +* Pamphile Roy + +* Pantelis Antonoudiou + +* Pearu Peterson +* Peter Andreas Entschev +* Peter Hawkins +* Pierre de Buyl +* Pieter Eendebak + +* Pradipta Ghosh + +* Rafael Cardoso Fernandes Sousa + +* Raghuveer Devulapalli +* Ralf Gommers +* Raphael Kruse +* Raúl Montón Pinillos +* Robert Kern +* Rohit Goswami +* Ross Barnowski +* Ruben Garcia + +* Sadie Louise Bartholomew + +* Saswat Das + +* Sayed Adel +* Sebastian Berg +* Serge Guelton +* Simon Surland Andersen + +* Siyabend Ürün + +* Somasree Majumder + +* Soumya + +* Stefan van der Walt +* Stefano Miccoli + +* Stephan Hoyer +* Stephen Worsley + +* Tania Allard +* Thomas Duvernay + +* Thomas Green + +* Thomas J. Fan +* Thomas Li + +* Tim Hoffmann +* Ting Sun + +* Tirth Patel +* Toshiki Kataoka +* Tyler Reddy +* Warren Weckesser +* Yang Hau +* Yoon, Jee Seok + + +Pull requests merged +==================== + +A total of 494 pull requests were merged for this release. + +* `#15006 <https://github.com/numpy/numpy/pull/15006>`__: ENH: add support for operator() in crackfortran. +* `#15844 <https://github.com/numpy/numpy/pull/15844>`__: ENH: add inline definition of access rights for Fortran types +* `#16810 <https://github.com/numpy/numpy/pull/16810>`__: MAINT: Remove subclass paths from scalar_value +* `#16830 <https://github.com/numpy/numpy/pull/16830>`__: MAINT: more python <3.6 cleanup +* `#16895 <https://github.com/numpy/numpy/pull/16895>`__: MAINT: extend delete single value optimization +* `#17709 <https://github.com/numpy/numpy/pull/17709>`__: BUG: Fix norm type promotion +* `#18343 <https://github.com/numpy/numpy/pull/18343>`__: DOC: document how to skip CI jobs +* `#18846 <https://github.com/numpy/numpy/pull/18846>`__: DOC: Improve documentation of default int type on windows +* `#19226 <https://github.com/numpy/numpy/pull/19226>`__: API: Fix structured dtype cast-safety, promotion, and comparison +* `#19345 <https://github.com/numpy/numpy/pull/19345>`__: ENH: Move ``ensure_dtype_nbo`` onto the DType as ``ensure_canonical`` +* `#19346 <https://github.com/numpy/numpy/pull/19346>`__: API: Fix ``np.result_type(structured_dtype)`` to "canonicalize" +* `#19581 <https://github.com/numpy/numpy/pull/19581>`__: DOC: Improve SIMD documentation(1/4) +* `#19756 <https://github.com/numpy/numpy/pull/19756>`__: DOC: Update front page of documentation with Sphinx-Panels +* `#19898 <https://github.com/numpy/numpy/pull/19898>`__: DOC: Fixed Refguide errors +* `#20020 <https://github.com/numpy/numpy/pull/20020>`__: ENH: add ndenumerate specialization for masked arrays +* `#20093 <https://github.com/numpy/numpy/pull/20093>`__: DOC: Created an indexing how-to +* `#20131 <https://github.com/numpy/numpy/pull/20131>`__: BUG: min/max is slow, re-implement using NEON (#17989) +* `#20133 <https://github.com/numpy/numpy/pull/20133>`__: ENH: Vectorize quicksort for 32-bit dtype using AVX-512 +* `#20140 <https://github.com/numpy/numpy/pull/20140>`__: DOC: Fix some target not found sphinx warnings. +* `#20147 <https://github.com/numpy/numpy/pull/20147>`__: DOC: updated docstring for binary file object +* `#20175 <https://github.com/numpy/numpy/pull/20175>`__: ENH: Optimize ``np.empty`` for scalar arguments +* `#20176 <https://github.com/numpy/numpy/pull/20176>`__: MAINT: Use intp output param viewable casts/methods +* `#20185 <https://github.com/numpy/numpy/pull/20185>`__: DOC: Added explanation document on interoperability +* `#20244 <https://github.com/numpy/numpy/pull/20244>`__: DOC: Clarify behavior of ``np.lib.scimath.sqrt`` apropos -0.0 +* `#20246 <https://github.com/numpy/numpy/pull/20246>`__: DOC: Merge doc strings of divide and true_divide. +* `#20285 <https://github.com/numpy/numpy/pull/20285>`__: ENH, SIMD: add new universal intrinsics for floor/rint +* `#20288 <https://github.com/numpy/numpy/pull/20288>`__: DOC: make some doctests in user,reference pass pytest +* `#20311 <https://github.com/numpy/numpy/pull/20311>`__: DOC: Windows and F2PY +* `#20363 <https://github.com/numpy/numpy/pull/20363>`__: SIMD: Replace SVML/ASM of tanh(f32, f64) with universal intrinsics +* `#20368 <https://github.com/numpy/numpy/pull/20368>`__: MAINT: Fix METH_NOARGS function signatures +* `#20380 <https://github.com/numpy/numpy/pull/20380>`__: DOC: random: Fix a comment and example in the multivariate_normal... +* `#20383 <https://github.com/numpy/numpy/pull/20383>`__: BLD: Try making 64-bit Windows wheels +* `#20387 <https://github.com/numpy/numpy/pull/20387>`__: REL: Prepare main for NumPy 1.23.0 development +* `#20388 <https://github.com/numpy/numpy/pull/20388>`__: Update ARM cpu_asimdfhm.c check +* `#20389 <https://github.com/numpy/numpy/pull/20389>`__: MAINT: Raise different type of errors +* `#20393 <https://github.com/numpy/numpy/pull/20393>`__: CI: CircleCI: Install numpy after processing doc_requirements.txt +* `#20394 <https://github.com/numpy/numpy/pull/20394>`__: DEP: remove allocation_tracking, deprecate PyDataMem_SetEventHook +* `#20395 <https://github.com/numpy/numpy/pull/20395>`__: ENH: provide a convenience function to replace npy_load_module +* `#20396 <https://github.com/numpy/numpy/pull/20396>`__: DOC: np.fromfunction documentation not clear +* `#20397 <https://github.com/numpy/numpy/pull/20397>`__: SIMD: replace raw AVX512 of floor/trunc/rint with universal intrinsics +* `#20398 <https://github.com/numpy/numpy/pull/20398>`__: BLD: Fix Macos Builds [wheel build] +* `#20399 <https://github.com/numpy/numpy/pull/20399>`__: DOC: Docstring improvements in the context of np.shape +* `#20403 <https://github.com/numpy/numpy/pull/20403>`__: CI: CircleCI: Install numpy after processing doc_requirements.txt +* `#20404 <https://github.com/numpy/numpy/pull/20404>`__: BUG: Clear errors correctly in F2PY conversions +* `#20405 <https://github.com/numpy/numpy/pull/20405>`__: BUG, SIMD: Fix ``exp`` FP stack overflow when ``AVX512_SKX`` is enabled +* `#20407 <https://github.com/numpy/numpy/pull/20407>`__: DOC: Update axis parameter for np.ma.{min,max} +* `#20409 <https://github.com/numpy/numpy/pull/20409>`__: MAINT: import setuptools before distutils in one ``np.random``... +* `#20412 <https://github.com/numpy/numpy/pull/20412>`__: MAINT: Fix METH_NOARGS function signatures (#20368) +* `#20413 <https://github.com/numpy/numpy/pull/20413>`__: DOC: correct the versionadded number for ``f2py.get_include`` +* `#20414 <https://github.com/numpy/numpy/pull/20414>`__: DEP: remove deprecated ``alen`` and ``asscalar`` functions +* `#20416 <https://github.com/numpy/numpy/pull/20416>`__: ENH: Add ARM Compiler with ARM Performance Library support +* `#20417 <https://github.com/numpy/numpy/pull/20417>`__: BLD: Add macOS arm64 wheels [wheel build] +* `#20422 <https://github.com/numpy/numpy/pull/20422>`__: BUG: Restore support for i386 and PowerPC (OS X) +* `#20427 <https://github.com/numpy/numpy/pull/20427>`__: MAINT: Fix longdouble precision check in test_umath.py +* `#20432 <https://github.com/numpy/numpy/pull/20432>`__: ENH: Add annotations for ``np.emath`` +* `#20433 <https://github.com/numpy/numpy/pull/20433>`__: BUG: Fix an incorrect protocol used in ``np.lib.shape_base`` +* `#20435 <https://github.com/numpy/numpy/pull/20435>`__: DOC: nicer CMake example in the f2py docs +* `#20437 <https://github.com/numpy/numpy/pull/20437>`__: DOC: Fix a typo in docstring of MT19937 +* `#20443 <https://github.com/numpy/numpy/pull/20443>`__: DOC: get scikit-build example working +* `#20446 <https://github.com/numpy/numpy/pull/20446>`__: BUG: Fixed output variable overriding in numpy.info() +* `#20447 <https://github.com/numpy/numpy/pull/20447>`__: DOC: use FindPython instead of FindPython3 +* `#20452 <https://github.com/numpy/numpy/pull/20452>`__: MAINT: Update the required setuptools version. +* `#20457 <https://github.com/numpy/numpy/pull/20457>`__: TST: remove obsolete TestF77Mismatch +* `#20468 <https://github.com/numpy/numpy/pull/20468>`__: BUG: Fix two overload-related problems +* `#20470 <https://github.com/numpy/numpy/pull/20470>`__: ENH: Add dtype-typing support to ``np.core.function_base`` +* `#20471 <https://github.com/numpy/numpy/pull/20471>`__: Rename _operand_flag_tests.c.src into numpy/core/src/umath/_operand_f… +* `#20478 <https://github.com/numpy/numpy/pull/20478>`__: TST,MAINT: F2PY test typo +* `#20479 <https://github.com/numpy/numpy/pull/20479>`__: TST,STY: Clean up F2PY tests for pathlib.Path +* `#20482 <https://github.com/numpy/numpy/pull/20482>`__: BUG: Fix tensorsolve for 0-sized input +* `#20484 <https://github.com/numpy/numpy/pull/20484>`__: BUG: Fix reduce promotion with out argument +* `#20486 <https://github.com/numpy/numpy/pull/20486>`__: MAINT: update wheel to version that supports python3.10 +* `#20489 <https://github.com/numpy/numpy/pull/20489>`__: MAINT: Translate binsearch.c.src to C++ using templates. +* `#20490 <https://github.com/numpy/numpy/pull/20490>`__: BUG: Protect divide by 0 in multinomial distribution. +* `#20491 <https://github.com/numpy/numpy/pull/20491>`__: TEST: use pypy3.8-v7.3.7 final versions +* `#20499 <https://github.com/numpy/numpy/pull/20499>`__: BUG: Fix the .T attribute in the array_api namespace +* `#20500 <https://github.com/numpy/numpy/pull/20500>`__: ENH: add ndmin to ``genfromtxt`` behaving the same as ``loadtxt`` +* `#20505 <https://github.com/numpy/numpy/pull/20505>`__: BUG: fix ``ma.average`` not working well with ``nan`` weights +* `#20509 <https://github.com/numpy/numpy/pull/20509>`__: DOC: Adds valgrind to the test command +* `#20515 <https://github.com/numpy/numpy/pull/20515>`__: ENH: Generate the docstrings of umath into a separate C header +* `#20516 <https://github.com/numpy/numpy/pull/20516>`__: DOC: Add more details on F2PY output conditions +* `#20517 <https://github.com/numpy/numpy/pull/20517>`__: MAINT,TST: Refactor F2PY testsuite +* `#20518 <https://github.com/numpy/numpy/pull/20518>`__: PERF: Fix performance bug in ufunc dispatching cache +* `#20521 <https://github.com/numpy/numpy/pull/20521>`__: MAINT: Pin OS versions when building wheels [wheel build] +* `#20524 <https://github.com/numpy/numpy/pull/20524>`__: CI: make sure CI stays on VS2019 unless changed explicitly +* `#20527 <https://github.com/numpy/numpy/pull/20527>`__: ENH: Add __array__ to the array_api Array object +* `#20528 <https://github.com/numpy/numpy/pull/20528>`__: BLD: Add PyPy wheels [wheel build] +* `#20533 <https://github.com/numpy/numpy/pull/20533>`__: BUG: Fix handling of the dtype parameter to numpy.array_api.prod() +* `#20547 <https://github.com/numpy/numpy/pull/20547>`__: REV: Revert adding a default ufunc promoter +* `#20552 <https://github.com/numpy/numpy/pull/20552>`__: ENH: Extending CPU feature detection framework to support IBM... +* `#20553 <https://github.com/numpy/numpy/pull/20553>`__: BLD: Use the new hypotl on Cygwin, rather than defaulting to... +* `#20556 <https://github.com/numpy/numpy/pull/20556>`__: DOC: Update links to mailing list on python.org +* `#20558 <https://github.com/numpy/numpy/pull/20558>`__: TST: move get_glibc_version to np.testing; skip 2 more tests... +* `#20559 <https://github.com/numpy/numpy/pull/20559>`__: DOC: Refactoring f2py user guide +* `#20563 <https://github.com/numpy/numpy/pull/20563>`__: BUG: Fix small issues found using valgrind +* `#20565 <https://github.com/numpy/numpy/pull/20565>`__: REF: Clean up wheels workflow [wheel build] +* `#20569 <https://github.com/numpy/numpy/pull/20569>`__: BUG: Fix sorting of int8/int16 +* `#20571 <https://github.com/numpy/numpy/pull/20571>`__: DOC: fix typo +* `#20572 <https://github.com/numpy/numpy/pull/20572>`__: DOC: Adds link to NEP 43 from NEP 41 +* `#20580 <https://github.com/numpy/numpy/pull/20580>`__: ENH: Move ``loadtxt`` to C for much better speed +* `#20583 <https://github.com/numpy/numpy/pull/20583>`__: BUG: Fix issues (mainly) found using pytest-leaks +* `#20587 <https://github.com/numpy/numpy/pull/20587>`__: MAINT: Fix two minor typing-related problems +* `#20588 <https://github.com/numpy/numpy/pull/20588>`__: BUG, DIST: fix normalize IBMZ features flags +* `#20589 <https://github.com/numpy/numpy/pull/20589>`__: DEP: remove NPY_ARRAY_UPDATEIFCOPY, deprecated in 1.14 +* `#20590 <https://github.com/numpy/numpy/pull/20590>`__: BUG: Fix leaks found using pytest-leaks +* `#20591 <https://github.com/numpy/numpy/pull/20591>`__: removed two redundant '\\' typos +* `#20592 <https://github.com/numpy/numpy/pull/20592>`__: BUG: Reject buffers with suboffsets +* `#20593 <https://github.com/numpy/numpy/pull/20593>`__: MAINT: Check for buffer interface support rather than try/except +* `#20594 <https://github.com/numpy/numpy/pull/20594>`__: BUG: Fix setstate logic for empty arrays +* `#20595 <https://github.com/numpy/numpy/pull/20595>`__: BUG: Fix PyInit__umath_linalg type +* `#20604 <https://github.com/numpy/numpy/pull/20604>`__: DEV: add a warningfilter to fix pytest workflow. +* `#20607 <https://github.com/numpy/numpy/pull/20607>`__: BUG: Protect kahan_sum from empty arrays +* `#20611 <https://github.com/numpy/numpy/pull/20611>`__: TST: Bump mypy: 0.910 -> 0.920 +* `#20616 <https://github.com/numpy/numpy/pull/20616>`__: MAINT: Help boost::python libraries at least not crash +* `#20621 <https://github.com/numpy/numpy/pull/20621>`__: BUG: random: Check 'writeable' flag in 'shuffle' and 'permuted'. +* `#20622 <https://github.com/numpy/numpy/pull/20622>`__: BLD: Add Windows 32-bit wheels +* `#20624 <https://github.com/numpy/numpy/pull/20624>`__: BUILD: pin to cython 0.29.24 to hide PyPy3.8 bug +* `#20628 <https://github.com/numpy/numpy/pull/20628>`__: REL: Update main after 1.21.5 release. +* `#20629 <https://github.com/numpy/numpy/pull/20629>`__: DOC: Refer to NumPy, not pandas, in main page +* `#20630 <https://github.com/numpy/numpy/pull/20630>`__: BUG: f2py: Simplify creation of an exception message. +* `#20640 <https://github.com/numpy/numpy/pull/20640>`__: BUG: Support env argument in CCompiler.spawn +* `#20641 <https://github.com/numpy/numpy/pull/20641>`__: PERF: Speed up check_constraint checks +* `#20643 <https://github.com/numpy/numpy/pull/20643>`__: PERF: Optimize array check for bounded 0,1 values +* `#20646 <https://github.com/numpy/numpy/pull/20646>`__: DOC: add np.iterable to reference guide +* `#20647 <https://github.com/numpy/numpy/pull/20647>`__: DOC: Add PyArray_FailUnlessWriteable to the online C-API docs. +* `#20648 <https://github.com/numpy/numpy/pull/20648>`__: DOC: Modify SVGs to be visible on Chrome +* `#20652 <https://github.com/numpy/numpy/pull/20652>`__: STY: Use PEP 585 and 604 syntaxes throughout the .pyi stub files +* `#20653 <https://github.com/numpy/numpy/pull/20653>`__: DEV: Add ``TYP``, a standard acronym for static typing +* `#20654 <https://github.com/numpy/numpy/pull/20654>`__: CI: Find cygwin test failures +* `#20660 <https://github.com/numpy/numpy/pull/20660>`__: MAINT: update OpenBLAS to 0.3.19 +* `#20663 <https://github.com/numpy/numpy/pull/20663>`__: TYP,TST: Bump mypy to 0.930 +* `#20666 <https://github.com/numpy/numpy/pull/20666>`__: DOC: Add help string for F2PY +* `#20668 <https://github.com/numpy/numpy/pull/20668>`__: TST: Initialize f2py2e tests of the F2PY CLI +* `#20669 <https://github.com/numpy/numpy/pull/20669>`__: CI, TST: Run Cygwin CI with Netlib reference BLAS and re-enable... +* `#20672 <https://github.com/numpy/numpy/pull/20672>`__: DOC: add hypothesis test dependency in README and PyPI long-description +* `#20674 <https://github.com/numpy/numpy/pull/20674>`__: BUG: array interface PyCapsule reference +* `#20678 <https://github.com/numpy/numpy/pull/20678>`__: BUG: Remove trailing dec point in dragon4positional +* `#20683 <https://github.com/numpy/numpy/pull/20683>`__: DOC: Updated pointer spacing for consistency. +* `#20689 <https://github.com/numpy/numpy/pull/20689>`__: BUG: Added check for NULL data in ufuncs +* `#20691 <https://github.com/numpy/numpy/pull/20691>`__: DOC, ENH: Added pngs for svgs for pdf build +* `#20693 <https://github.com/numpy/numpy/pull/20693>`__: DOC: Replaced svgs with pngs in the Broadcasting doc +* `#20695 <https://github.com/numpy/numpy/pull/20695>`__: BLD: Add NPY_DISABLE_SVML env var to opt out of SVML +* `#20697 <https://github.com/numpy/numpy/pull/20697>`__: REL: Update main after 1.22.0 release. +* `#20698 <https://github.com/numpy/numpy/pull/20698>`__: DOC:Fixed the link on user-guide landing page +* `#20701 <https://github.com/numpy/numpy/pull/20701>`__: MAINT, DOC: Post 1.22.0 release fixes. +* `#20708 <https://github.com/numpy/numpy/pull/20708>`__: DOC: fix broken documentation references in mtrand.pyx +* `#20710 <https://github.com/numpy/numpy/pull/20710>`__: TYP: Allow ``ndindex`` to accept integer tuples +* `#20712 <https://github.com/numpy/numpy/pull/20712>`__: BUG: Restore vc141 support +* `#20713 <https://github.com/numpy/numpy/pull/20713>`__: DOC: Add Code of Conduct to README.md +* `#20719 <https://github.com/numpy/numpy/pull/20719>`__: TYP: change type annotation for ``__array_namespace__`` to ModuleType +* `#20720 <https://github.com/numpy/numpy/pull/20720>`__: TYP: add a few type annotations to ``numpy.array_api.Array`` +* `#20721 <https://github.com/numpy/numpy/pull/20721>`__: BUG: Fix array dimensions solver for multidimensional arguments... +* `#20722 <https://github.com/numpy/numpy/pull/20722>`__: ENH: Removed requirement for C-contiguity when changing to dtype... +* `#20727 <https://github.com/numpy/numpy/pull/20727>`__: DOC: Update README.md mainly to include link to website +* `#20729 <https://github.com/numpy/numpy/pull/20729>`__: BUG: Relax dtype identity check in reductions +* `#20730 <https://github.com/numpy/numpy/pull/20730>`__: DOC: Document that dtype, strides, shape attributes should not... +* `#20731 <https://github.com/numpy/numpy/pull/20731>`__: DOC: fix OpenBLAS version in release note +* `#20732 <https://github.com/numpy/numpy/pull/20732>`__: MAINT: Translate timsort.c.src to C++ using templates. +* `#20738 <https://github.com/numpy/numpy/pull/20738>`__: ENH: fix a typo in the example trigger for wheels +* `#20740 <https://github.com/numpy/numpy/pull/20740>`__: Update teams URL +* `#20741 <https://github.com/numpy/numpy/pull/20741>`__: DOC: add instructions for cross compilation +* `#20745 <https://github.com/numpy/numpy/pull/20745>`__: ENH: add hook and test for PyInstaller. +* `#20747 <https://github.com/numpy/numpy/pull/20747>`__: BLD: Upload wheel artifacts separately [wheel build] +* `#20750 <https://github.com/numpy/numpy/pull/20750>`__: TYP: Allow time manipulation functions to accept ``date`` and ``timedelta``... +* `#20754 <https://github.com/numpy/numpy/pull/20754>`__: MAINT: Relax asserts to match relaxed reducelike resolution behaviour +* `#20758 <https://github.com/numpy/numpy/pull/20758>`__: DOC: Capitalization and missing word in docs +* `#20759 <https://github.com/numpy/numpy/pull/20759>`__: MAINT: Raise RuntimeError if setuptools version is too recent. +* `#20762 <https://github.com/numpy/numpy/pull/20762>`__: BUG: Allow integer inputs for pow-related functions in ``array_api`` +* `#20766 <https://github.com/numpy/numpy/pull/20766>`__: ENH: Make ndarray.__array_finalize__ a callable no-op +* `#20773 <https://github.com/numpy/numpy/pull/20773>`__: BUG: method without self argument should be static +* `#20774 <https://github.com/numpy/numpy/pull/20774>`__: DOC: explicitly define numpy.datetime64 semantics +* `#20776 <https://github.com/numpy/numpy/pull/20776>`__: DOC: fix remaining "easy" doctests errors +* `#20779 <https://github.com/numpy/numpy/pull/20779>`__: MAINT: removed duplicate 'int' type in ScalarType +* `#20783 <https://github.com/numpy/numpy/pull/20783>`__: DOC: Update Copyright to 2022 [License] +* `#20784 <https://github.com/numpy/numpy/pull/20784>`__: MAINT, DOC: fix new typos detected by codespell +* `#20786 <https://github.com/numpy/numpy/pull/20786>`__: BUG, DOC: Fixes SciPy docs build warnings +* `#20788 <https://github.com/numpy/numpy/pull/20788>`__: BUG: ``array_api.argsort(descending=True)`` respects relative... +* `#20789 <https://github.com/numpy/numpy/pull/20789>`__: DOC: git:// protocol deprecated by github. +* `#20791 <https://github.com/numpy/numpy/pull/20791>`__: BUG: Return correctly shaped inverse indices in ``array_api`` set... +* `#20792 <https://github.com/numpy/numpy/pull/20792>`__: TST: Bump mypy to 0.931 +* `#20793 <https://github.com/numpy/numpy/pull/20793>`__: BUG: Fix that reduce-likes honor out always (and live in the... +* `#20794 <https://github.com/numpy/numpy/pull/20794>`__: TYP: Type the NEP 35 ``like`` parameter via a ``__array_function__``... +* `#20810 <https://github.com/numpy/numpy/pull/20810>`__: DOC: Restore MaskedArray.hardmask documentation +* `#20811 <https://github.com/numpy/numpy/pull/20811>`__: MAINT, DOC: discard repeated words +* `#20813 <https://github.com/numpy/numpy/pull/20813>`__: MAINT: fix typo +* `#20816 <https://github.com/numpy/numpy/pull/20816>`__: DOC: discard repeated words in NEPs +* `#20818 <https://github.com/numpy/numpy/pull/20818>`__: BUG: Fix build of third-party extensions with Py_LIMITED_API +* `#20821 <https://github.com/numpy/numpy/pull/20821>`__: ENH: Add CPU feature detection for POWER10 (VSX4) +* `#20823 <https://github.com/numpy/numpy/pull/20823>`__: REL: Update main after 1.22.1 release. +* `#20827 <https://github.com/numpy/numpy/pull/20827>`__: TYP: Fix pyright being unable to infer the ``real`` and ``imag``... +* `#20828 <https://github.com/numpy/numpy/pull/20828>`__: MAINT: Translate heapsort.c.src to C++ using templates +* `#20829 <https://github.com/numpy/numpy/pull/20829>`__: MAINT: Translate mergesort.c.src to C++ using templates. +* `#20831 <https://github.com/numpy/numpy/pull/20831>`__: BUG: Avoid importing numpy.distutils on import numpy.testing +* `#20833 <https://github.com/numpy/numpy/pull/20833>`__: BUG: Fix comparator function signatures +* `#20834 <https://github.com/numpy/numpy/pull/20834>`__: DOC: Update ndarray.argmax + argmin documentation with keepdims... +* `#20835 <https://github.com/numpy/numpy/pull/20835>`__: DEP: Removed deprecated error clearing +* `#20840 <https://github.com/numpy/numpy/pull/20840>`__: MAINT: Translate selection.c.src to C++ using templates. +* `#20846 <https://github.com/numpy/numpy/pull/20846>`__: ENH, SIMD: improve argmax/argmin performance +* `#20847 <https://github.com/numpy/numpy/pull/20847>`__: MAINT: remove outdated mingw32 fseek support +* `#20851 <https://github.com/numpy/numpy/pull/20851>`__: DOC: Fix typo in meshgrid example +* `#20852 <https://github.com/numpy/numpy/pull/20852>`__: MAINT: Fix a typo in numpy/f2py/capi_maps.py +* `#20854 <https://github.com/numpy/numpy/pull/20854>`__: DEV: Update dependencies and Docker image +* `#20857 <https://github.com/numpy/numpy/pull/20857>`__: BUG: Fix pre-builds in Gitpod +* `#20858 <https://github.com/numpy/numpy/pull/20858>`__: TYP: Relax the return-type of ``np.vectorize`` +* `#20861 <https://github.com/numpy/numpy/pull/20861>`__: DOC: fix formatting of mean example +* `#20862 <https://github.com/numpy/numpy/pull/20862>`__: Fix typo in numpy/lib/polynomial.py +* `#20865 <https://github.com/numpy/numpy/pull/20865>`__: MAINT: Fix inconsistent PyPI casing +* `#20866 <https://github.com/numpy/numpy/pull/20866>`__: ENH: Add changes that allow NumPy to compile with clang-cl +* `#20867 <https://github.com/numpy/numpy/pull/20867>`__: DOC: Cosmetic docstring fix for numpydoc. +* `#20868 <https://github.com/numpy/numpy/pull/20868>`__: BUG: Gitpod Remove lock file --unshallow +* `#20869 <https://github.com/numpy/numpy/pull/20869>`__: DOC: random: Fix spelling of 'precision'. +* `#20872 <https://github.com/numpy/numpy/pull/20872>`__: BUG: Loss of precision in longdouble min +* `#20874 <https://github.com/numpy/numpy/pull/20874>`__: BUG: mtrand cannot be imported on Cygwin +* `#20875 <https://github.com/numpy/numpy/pull/20875>`__: DEP: deprecate ``numpy.distutils``, and add a migration guide +* `#20876 <https://github.com/numpy/numpy/pull/20876>`__: MAINT, DOC: Fixes minor formatting issue related to nested inline... +* `#20878 <https://github.com/numpy/numpy/pull/20878>`__: DOC,TST: Fix Pandas code example +* `#20881 <https://github.com/numpy/numpy/pull/20881>`__: BUG: fix f2py's define for threading when building with Mingw +* `#20883 <https://github.com/numpy/numpy/pull/20883>`__: BUG: Fix ``np.array_api.can_cast()`` by not relying on ``np.can_cast()`` +* `#20884 <https://github.com/numpy/numpy/pull/20884>`__: MAINT: Minor cleanup to F2PY +* `#20885 <https://github.com/numpy/numpy/pull/20885>`__: TYP,ENH: Improve typing with the help of ``ParamSpec`` +* `#20886 <https://github.com/numpy/numpy/pull/20886>`__: BUG: distutils: fix building mixed C/Fortran extensions +* `#20887 <https://github.com/numpy/numpy/pull/20887>`__: TYP,MAINT: Add aliases for commonly used unions +* `#20890 <https://github.com/numpy/numpy/pull/20890>`__: BUILD: Upload wheels to anaconda,org +* `#20897 <https://github.com/numpy/numpy/pull/20897>`__: MAINT: Translate quicksort.c.src to C++ using templates. +* `#20900 <https://github.com/numpy/numpy/pull/20900>`__: TYP,ENH: Add annotations for ``np.lib.mixins`` +* `#20902 <https://github.com/numpy/numpy/pull/20902>`__: TYP,ENH: Add dtype-typing support to ``np.core.fromnumeric`` (part... +* `#20904 <https://github.com/numpy/numpy/pull/20904>`__: ENH,BUG: Expand the experimental DType API and fix small exposed... +* `#20911 <https://github.com/numpy/numpy/pull/20911>`__: BUG: Fix the return type of random_float_fill +* `#20916 <https://github.com/numpy/numpy/pull/20916>`__: TYP, MAINT: Add annotations for ``flatiter.__setitem__`` +* `#20917 <https://github.com/numpy/numpy/pull/20917>`__: DOC: fix np.ma.flatnotmasked_contiguous docstring +* `#20918 <https://github.com/numpy/numpy/pull/20918>`__: MAINT, TYP: Added missing where typehints in fromnumeric.pyi +* `#20920 <https://github.com/numpy/numpy/pull/20920>`__: DEP: Deprecate use of ``axis=MAXDIMS`` instead of ``axis=None`` +* `#20927 <https://github.com/numpy/numpy/pull/20927>`__: DOC: lib/io.py was renamed to lib/npyio.py +* `#20931 <https://github.com/numpy/numpy/pull/20931>`__: BUG: Fix missing intrinsics for windows/arm64 target +* `#20934 <https://github.com/numpy/numpy/pull/20934>`__: BUG: Fix build_ext interaction with non-numpy extensions +* `#20940 <https://github.com/numpy/numpy/pull/20940>`__: MAINT: f2py: don't generate code that triggers ``-Wsometimes-uninitialized`` +* `#20944 <https://github.com/numpy/numpy/pull/20944>`__: DOC: improper doc syntax (markdown and imbalanced ticks). +* `#20946 <https://github.com/numpy/numpy/pull/20946>`__: MAINT: Fix typo in setup.py +* `#20948 <https://github.com/numpy/numpy/pull/20948>`__: MAINT, DOC: NEP link update +* `#20950 <https://github.com/numpy/numpy/pull/20950>`__: Fix broken link in nep-0046-sponsorship-guidelines.rst +* `#20955 <https://github.com/numpy/numpy/pull/20955>`__: BUG: Fix incorrect return type in reduce without initial value +* `#20956 <https://github.com/numpy/numpy/pull/20956>`__: DOC: Improve NEP page layout with nested toctrees +* `#20960 <https://github.com/numpy/numpy/pull/20960>`__: ENH: review return values for PyArray_DescrNew +* `#20963 <https://github.com/numpy/numpy/pull/20963>`__: MAINT: be more tolerant of setuptools>=60 +* `#20966 <https://github.com/numpy/numpy/pull/20966>`__: DOC: update python minimal version to build from source +* `#20967 <https://github.com/numpy/numpy/pull/20967>`__: MAINT: Update to numpydoc v1.2 +* `#20968 <https://github.com/numpy/numpy/pull/20968>`__: MAINT: Translate npy_partition.h.src to C++ using templates. +* `#20972 <https://github.com/numpy/numpy/pull/20972>`__: DOC: Add warning about differences between range and arange +* `#20973 <https://github.com/numpy/numpy/pull/20973>`__: DOC: switch Python intersphinx link from dev to stable. +* `#20974 <https://github.com/numpy/numpy/pull/20974>`__: DOC: Include special case in ``hsplit`` doc +* `#20975 <https://github.com/numpy/numpy/pull/20975>`__: MAINT: refactor NonNull in API functions +* `#20976 <https://github.com/numpy/numpy/pull/20976>`__: ENH,BENCH: Optimize floor_divide for VSX4/Power10 +* `#20987 <https://github.com/numpy/numpy/pull/20987>`__: BLD: Try adding aarch64 wheels [wheel build] +* `#20990 <https://github.com/numpy/numpy/pull/20990>`__: MAINT: Further small return value validation fixes +* `#20991 <https://github.com/numpy/numpy/pull/20991>`__: ENH: Use SVML for f64 exp and log +* `#20993 <https://github.com/numpy/numpy/pull/20993>`__: ENH: Allow object and subarray dtypes in fromiter +* `#20994 <https://github.com/numpy/numpy/pull/20994>`__: REL: Update main after 1.22.2 release. +* `#20996 <https://github.com/numpy/numpy/pull/20996>`__: MAINT: use brackets in github action syntax +* `#20999 <https://github.com/numpy/numpy/pull/20999>`__: DOC: Remove mention of deleted subpackages in numpy docstring +* `#21000 <https://github.com/numpy/numpy/pull/21000>`__: MAINT: Replace LooseVersion by _pep440. +* `#21001 <https://github.com/numpy/numpy/pull/21001>`__: ENH: help compilers to auto-vectorize reduction operators +* `#21003 <https://github.com/numpy/numpy/pull/21003>`__: ENH: Suppress over-/underflow RuntimeWarning in assert_array_equal +* `#21005 <https://github.com/numpy/numpy/pull/21005>`__: BUG: Add parameter check to negative_binomial +* `#21010 <https://github.com/numpy/numpy/pull/21010>`__: MAINT: Fix warning message for deprecated keyword +* `#21015 <https://github.com/numpy/numpy/pull/21015>`__: DOC: Added note about possible arange signatures +* `#21016 <https://github.com/numpy/numpy/pull/21016>`__: MAINT, DOC: Fix SciPy intersphinx link +* `#21020 <https://github.com/numpy/numpy/pull/21020>`__: DOC: imbalanced backticks +* `#21021 <https://github.com/numpy/numpy/pull/21021>`__: TYP,ENH: Add dtype-typing support to ``fromnumeric`` (part 2) +* `#21024 <https://github.com/numpy/numpy/pull/21024>`__: API: Disallow strings in logical ufuncs +* `#21025 <https://github.com/numpy/numpy/pull/21025>`__: MAINT: Use C++ for tokenizer unicode-kind templating +* `#21027 <https://github.com/numpy/numpy/pull/21027>`__: BUG: use ``concurrent.futures.ThreadPoolExecutor`` in distutils... +* `#21029 <https://github.com/numpy/numpy/pull/21029>`__: DEP: Remove support for non-tuple nd-indices. +* `#21030 <https://github.com/numpy/numpy/pull/21030>`__: DOC: change fill_value of full_like from scalar to array_like +* `#21031 <https://github.com/numpy/numpy/pull/21031>`__: MAINT, STY: Style fixes to quicksort.cpp +* `#21032 <https://github.com/numpy/numpy/pull/21032>`__: DOC: fix sphinx errors due to np.emath references +* `#21035 <https://github.com/numpy/numpy/pull/21035>`__: BUILD: remove condition on upload step +* `#21037 <https://github.com/numpy/numpy/pull/21037>`__: DOC: Consistency of :: syntax. +* `#21039 <https://github.com/numpy/numpy/pull/21039>`__: MAINT: Remove the RELAXED_STRIDES_CHECKING env variable +* `#21040 <https://github.com/numpy/numpy/pull/21040>`__: DOC: "See Also" should not have backticks. +* `#21042 <https://github.com/numpy/numpy/pull/21042>`__: MAINT, STY: Style fixups. +* `#21043 <https://github.com/numpy/numpy/pull/21043>`__: BUILD: simplify upload step +* `#21045 <https://github.com/numpy/numpy/pull/21045>`__: BUILD: change syntax to use env variable +* `#21046 <https://github.com/numpy/numpy/pull/21046>`__: MAINT: Use "3.10" instead of "3.10-dev" on travis. +* `#21049 <https://github.com/numpy/numpy/pull/21049>`__: use repo secrets for uploading +* `#21050 <https://github.com/numpy/numpy/pull/21050>`__: BUILD: tweak upload to use python3, less verbose +* `#21053 <https://github.com/numpy/numpy/pull/21053>`__: BUILD: make sure a python3 is on the path +* `#21054 <https://github.com/numpy/numpy/pull/21054>`__: BUG: (loadtxt) Ignore last empty field when ``delimiter=None`` +* `#21060 <https://github.com/numpy/numpy/pull/21060>`__: TYP: Add dtype-typing support to ``fromnumeric`` part 3 +* `#21061 <https://github.com/numpy/numpy/pull/21061>`__: BLD,ENH: Add vsx3 and vsx4 as targets when building cos/sin and... +* `#21064 <https://github.com/numpy/numpy/pull/21064>`__: DOC: Update arctan2 docstring based on doctest output +* `#21067 <https://github.com/numpy/numpy/pull/21067>`__: BUG: Fix unpickling an empty ndarray with a non-zero dimension +* `#21068 <https://github.com/numpy/numpy/pull/21068>`__: DOC: Fix spelling and grammar in documentation for quantile(). +* `#21071 <https://github.com/numpy/numpy/pull/21071>`__: BUG: Ensure equality/identity comparison with ``__array_function__`` +* `#21074 <https://github.com/numpy/numpy/pull/21074>`__: BUG: Replace ``ssize_t`` with ``size_t`` in tokenize.cpp +* `#21077 <https://github.com/numpy/numpy/pull/21077>`__: TYP,MAINT: Remove inconsistencies between ``fromnumeric`` functions... +* `#21082 <https://github.com/numpy/numpy/pull/21082>`__: DOC: clarify the return value of linalg.cholesky +* `#21085 <https://github.com/numpy/numpy/pull/21085>`__: MAINT: point to html docs on distutils migration in deprecation... +* `#21086 <https://github.com/numpy/numpy/pull/21086>`__: DEV: fix ``python runtests.py --bench-compare`` +* `#21087 <https://github.com/numpy/numpy/pull/21087>`__: DOC: update docs in site.cfg.example +* `#21088 <https://github.com/numpy/numpy/pull/21088>`__: DEV: add distutils deprecation warning filter to pytest conf +* `#21089 <https://github.com/numpy/numpy/pull/21089>`__: BUILD: if travis build is triggered manually, then upload wheels +* `#21090 <https://github.com/numpy/numpy/pull/21090>`__: BUILD: change cibuildwheel output directory on travis [ci skip] +* `#21095 <https://github.com/numpy/numpy/pull/21095>`__: BLD: Make a sdist [wheel build] +* `#21097 <https://github.com/numpy/numpy/pull/21097>`__: MAINT: update cython, pypy for cython0.29.28 and pypy v7.3.8... +* `#21099 <https://github.com/numpy/numpy/pull/21099>`__: MAINT: Translate x86-qsort.dispatch.c.src to C++ using templates. +* `#21100 <https://github.com/numpy/numpy/pull/21100>`__: BLD: Comment out broken macOS PyPy build [wheel build] +* `#21102 <https://github.com/numpy/numpy/pull/21102>`__: TYP,MAINT: Explicitly allow sequences of array-likes in ``np.concatenate`` +* `#21107 <https://github.com/numpy/numpy/pull/21107>`__: BLD: Run wheel builders on labeled pull requests +* `#21108 <https://github.com/numpy/numpy/pull/21108>`__: TYP, ENH: Mark non-subclassable classes as ``final`` +* `#21109 <https://github.com/numpy/numpy/pull/21109>`__: MAINT: Fix incorrect signature in readtext header file +* `#21110 <https://github.com/numpy/numpy/pull/21110>`__: CI: Improve concurrency to cancel running jobs on PR update +* `#21111 <https://github.com/numpy/numpy/pull/21111>`__: TYP, MAINT: Relax the ``obj`` type in ``__array_finalize__`` +* `#21113 <https://github.com/numpy/numpy/pull/21113>`__: BUG: Fix numba DUFuncs added loops getting picked up +* `#21118 <https://github.com/numpy/numpy/pull/21118>`__: DOC: improve documentation of singular value decomposition +* `#21119 <https://github.com/numpy/numpy/pull/21119>`__: BUG, ENH: np._from_dlpack: export correct device information +* `#21121 <https://github.com/numpy/numpy/pull/21121>`__: MAINT,TST: np._from_dlpack: add more test + small memory optimization +* `#21124 <https://github.com/numpy/numpy/pull/21124>`__: ENH,SIMD: Vectorize modulo/divide using the universal intrinsics... +* `#21125 <https://github.com/numpy/numpy/pull/21125>`__: BLD: bump cibuildwheel 2.3.0 → 2.3.1 on GHA [wheel build] +* `#21127 <https://github.com/numpy/numpy/pull/21127>`__: BLD,DOC: skip broken ipython 8.1.0 +* `#21128 <https://github.com/numpy/numpy/pull/21128>`__: BLD: move cibuildwheel configuration to ``pyproject.toml`` +* `#21130 <https://github.com/numpy/numpy/pull/21130>`__: ENH: improve the speed of numpy.where using a branchless code +* `#21132 <https://github.com/numpy/numpy/pull/21132>`__: BUG,ENH: np._from_dlpack: export arrays with any-strided size-1... +* `#21133 <https://github.com/numpy/numpy/pull/21133>`__: DOC: Note interop from "subclassing" docs and explain when to... +* `#21144 <https://github.com/numpy/numpy/pull/21144>`__: DOC: Change recommendation away from pinning numpy+3 +* `#21145 <https://github.com/numpy/numpy/pull/21145>`__: MAINT, DOC: make np._from_dlpack public +* `#21146 <https://github.com/numpy/numpy/pull/21146>`__: BUG: assign all tuple items before using it for PyPy +* `#21149 <https://github.com/numpy/numpy/pull/21149>`__: DOC: Update linalg.qr docstring with numerically stable example +* `#21150 <https://github.com/numpy/numpy/pull/21150>`__: DOC: Fix syntax highlighting for numpy.flatnonzero +* `#21151 <https://github.com/numpy/numpy/pull/21151>`__: ENH: Add 'ulong' to sctypeDict +* `#21154 <https://github.com/numpy/numpy/pull/21154>`__: ENH, BLD: Fix math feature detection for wasm +* `#21155 <https://github.com/numpy/numpy/pull/21155>`__: DOC: document uploads to ananconda.org +* `#21157 <https://github.com/numpy/numpy/pull/21157>`__: DOC: fix documentation for typedescr argument of PyArray_AsCArray +* `#21167 <https://github.com/numpy/numpy/pull/21167>`__: DOC: Add "pip install -r test_requirements.txt" +* `#21170 <https://github.com/numpy/numpy/pull/21170>`__: REL: Update main after 1.22.3 release. +* `#21178 <https://github.com/numpy/numpy/pull/21178>`__: MAINT: Move can-cast table to a custom header file +* `#21180 <https://github.com/numpy/numpy/pull/21180>`__: TST: Bump mypy from 0.931 to 0.940 +* `#21185 <https://github.com/numpy/numpy/pull/21185>`__: TYP, BUG: Fix ``np.lib.stride_tricks`` re-exported under wrong... +* `#21186 <https://github.com/numpy/numpy/pull/21186>`__: MAINT: update NEP 29 +* `#21187 <https://github.com/numpy/numpy/pull/21187>`__: ENH: F2PY build output determinism +* `#21188 <https://github.com/numpy/numpy/pull/21188>`__: MAINT,ENH: Rewrite scalar math logic +* `#21189 <https://github.com/numpy/numpy/pull/21189>`__: DEV: Remove deprecated "python.pythonPath" +* `#21193 <https://github.com/numpy/numpy/pull/21193>`__: DOC: Remove the confusing "unless not" in numpy/core/fromnumeric.py +* `#21201 <https://github.com/numpy/numpy/pull/21201>`__: DOC: typo corrected in numpy.argpartition +* `#21202 <https://github.com/numpy/numpy/pull/21202>`__: DOC: fix outdated description of unicode +* `#21205 <https://github.com/numpy/numpy/pull/21205>`__: BUG: f2py cannot read in customised f2cmap file; fix #21204 +* `#21206 <https://github.com/numpy/numpy/pull/21206>`__: MAINT: fix typo in NEP 29 +* `#21207 <https://github.com/numpy/numpy/pull/21207>`__: MAINT: remove maint from triggering wheel build, add env to sdist +* `#21216 <https://github.com/numpy/numpy/pull/21216>`__: MAINT: Split ``numpy.typing`` into a public and private component +* `#21218 <https://github.com/numpy/numpy/pull/21218>`__: BUG: Use -0. as initial value for summation (internal only) +* `#21226 <https://github.com/numpy/numpy/pull/21226>`__: DOC: misc fixes +* `#21227 <https://github.com/numpy/numpy/pull/21227>`__: MAINT: Translate numpy/linalg/umath_linalg.c.src to C++ using... +* `#21231 <https://github.com/numpy/numpy/pull/21231>`__: BUG: Catch error if array-priority is not float compatible +* `#21232 <https://github.com/numpy/numpy/pull/21232>`__: BUG: Fixes ``ValueError`` in ``np.kron`` +* `#21238 <https://github.com/numpy/numpy/pull/21238>`__: BLD: Fix upload script +* `#21241 <https://github.com/numpy/numpy/pull/21241>`__: MAINT: use doc_requirements.txt in azure build +* `#21244 <https://github.com/numpy/numpy/pull/21244>`__: TST: Bump mypy from 0.940 to 0.942 +* `#21247 <https://github.com/numpy/numpy/pull/21247>`__: DOC: directive fix (single instead of double backticks). +* `#21250 <https://github.com/numpy/numpy/pull/21250>`__: DEV: Fixed Un-responsive live-preview in gitpod. +* `#21251 <https://github.com/numpy/numpy/pull/21251>`__: DOC: fix data type of parameter shape +* `#21253 <https://github.com/numpy/numpy/pull/21253>`__: DOC: fix code sample for leg2poly +* `#21254 <https://github.com/numpy/numpy/pull/21254>`__: DOC: document automatic wheel building for a release +* `#21255 <https://github.com/numpy/numpy/pull/21255>`__: DOC: mention Gitpod as alternative to build numpy +* `#21256 <https://github.com/numpy/numpy/pull/21256>`__: BUG,ENH: Fix negative bounds for F2PY +* `#21260 <https://github.com/numpy/numpy/pull/21260>`__: DOC: Enumerate the differences between numpy and numpy.array_api +* `#21262 <https://github.com/numpy/numpy/pull/21262>`__: ENH: Masked Array support for ``np.kron`` +* `#21269 <https://github.com/numpy/numpy/pull/21269>`__: DOC: Improve "random.generator.shuffle" docs page +* `#21272 <https://github.com/numpy/numpy/pull/21272>`__: BUG: Fix typos +* `#21285 <https://github.com/numpy/numpy/pull/21285>`__: BLD: Bump cibuildwheel and enable more PyPy +* `#21286 <https://github.com/numpy/numpy/pull/21286>`__: DOC: double backticks and links +* `#21287 <https://github.com/numpy/numpy/pull/21287>`__: MAINT: Use C++ inline and include files in C++ files. +* `#21290 <https://github.com/numpy/numpy/pull/21290>`__: DOC: Improve documentation formatting +* `#21291 <https://github.com/numpy/numpy/pull/21291>`__: DOC: Add space after argument name +* `#21295 <https://github.com/numpy/numpy/pull/21295>`__: MAINT: Clean-up includes of auto-generated umath code +* `#21297 <https://github.com/numpy/numpy/pull/21297>`__: MAINT: Rename source files that were not using any template-preprocessing +* `#21303 <https://github.com/numpy/numpy/pull/21303>`__: MAINT: Edit logo size and logo position in README.md +* `#21306 <https://github.com/numpy/numpy/pull/21306>`__: ENH: Introduce numpy.core.setup_common.NPY_CXX_FLAGS +* `#21307 <https://github.com/numpy/numpy/pull/21307>`__: MAINT: bump versions in Github actions configuration to v3. +* `#21314 <https://github.com/numpy/numpy/pull/21314>`__: DOC: various spell checks and typo fixes +* `#21315 <https://github.com/numpy/numpy/pull/21315>`__: DOC: minor typo fix in numpy.random API docs +* `#21321 <https://github.com/numpy/numpy/pull/21321>`__: BUG: Stop using PyBytesObject.ob_shash deprecated in Python 3.11. +* `#21324 <https://github.com/numpy/numpy/pull/21324>`__: BUG: Make mmap handling safer in frombuffer +* `#21327 <https://github.com/numpy/numpy/pull/21327>`__: Small updates to the array_api docs +* `#21330 <https://github.com/numpy/numpy/pull/21330>`__: DOC: Add F2PY tests documentation +* `#21331 <https://github.com/numpy/numpy/pull/21331>`__: REL: Update main after 1.21.6 release. +* `#21345 <https://github.com/numpy/numpy/pull/21345>`__: TYP: Let ``ndarray`` fancy indexing always return an ``ndarray`` +* `#21347 <https://github.com/numpy/numpy/pull/21347>`__: MAINT: Fix failing simd and cygwin tests. +* `#21348 <https://github.com/numpy/numpy/pull/21348>`__: DEV: reverted misplaced install of "esbonio". +* `#21349 <https://github.com/numpy/numpy/pull/21349>`__: MAINT: Update setup-cygwin to v3 again. +* `#21352 <https://github.com/numpy/numpy/pull/21352>`__: Doc: Philox.jumped correct the formula +* `#21354 <https://github.com/numpy/numpy/pull/21354>`__: ENH: Improve ``np.kron`` performance +* `#21355 <https://github.com/numpy/numpy/pull/21355>`__: MAINT: Remove the reference to the “good first issue” label +* `#21356 <https://github.com/numpy/numpy/pull/21356>`__: DOC: Fix a typo in docstring of MT19937 +* `#21360 <https://github.com/numpy/numpy/pull/21360>`__: MAINT: Add compile flag to disable voltbl on MSVC 142 +* `#21366 <https://github.com/numpy/numpy/pull/21366>`__: BUG: fix compilation error for VS 141 and earlier +* `#21367 <https://github.com/numpy/numpy/pull/21367>`__: MAINT: Translate ieee754.c.src to C++ using templates. +* `#21368 <https://github.com/numpy/numpy/pull/21368>`__: MAINT: Fix failing Python 3.8 32-bit Windows test. +* `#21372 <https://github.com/numpy/numpy/pull/21372>`__: BUG: Allow legacy dtypes to cast to datetime again +* `#21377 <https://github.com/numpy/numpy/pull/21377>`__: API: Allow newaxis indexing for ``array_api`` arrays +* `#21381 <https://github.com/numpy/numpy/pull/21381>`__: DOC: Typesetting of math for np.correlate and np.convolve +* `#21382 <https://github.com/numpy/numpy/pull/21382>`__: DOC: non-orphan page, and casing. +* `#21384 <https://github.com/numpy/numpy/pull/21384>`__: BUG: Missing ``f`` prefix on f-strings fix +* `#21388 <https://github.com/numpy/numpy/pull/21388>`__: MAINT: be sure to match base and docker images +* `#21392 <https://github.com/numpy/numpy/pull/21392>`__: BUG: add linux guard per #21386 +* `#21394 <https://github.com/numpy/numpy/pull/21394>`__: PERF: Reduce overhead of np.linalg.norm for small arrays +* `#21400 <https://github.com/numpy/numpy/pull/21400>`__: DOC: Add missing entries in ``numpy.testing`` documentation +* `#21407 <https://github.com/numpy/numpy/pull/21407>`__: MAINT: Reduce f2py verbiage for valid parameters +* `#21410 <https://github.com/numpy/numpy/pull/21410>`__: DOC: Update set of allowed f2cmap types +* `#21411 <https://github.com/numpy/numpy/pull/21411>`__: MAINT: Remove ``f2py.f2py_testing`` without replacement +* `#21413 <https://github.com/numpy/numpy/pull/21413>`__: DOC: Secure PR template URLs [ci skip] +* `#21415 <https://github.com/numpy/numpy/pull/21415>`__: BUG: Fix handling of skip-empty-wrappers +* `#21417 <https://github.com/numpy/numpy/pull/21417>`__: MAINT: Update doc requirements +* `#21421 <https://github.com/numpy/numpy/pull/21421>`__: MAINT: Remove FPE helper code that is unnecessary on C99/C++11 +* `#21423 <https://github.com/numpy/numpy/pull/21423>`__: PERF: Improve performance of special attribute lookups +* `#21425 <https://github.com/numpy/numpy/pull/21425>`__: TEST: on PyPy, skip hanging slow test [wheel build] +* `#21426 <https://github.com/numpy/numpy/pull/21426>`__: DOC: Add version switcher to the documentation +* `#21430 <https://github.com/numpy/numpy/pull/21430>`__: TYP: Bump mypy to 0.950 +* `#21436 <https://github.com/numpy/numpy/pull/21436>`__: BUG: Fix segmentation fault +* `#21442 <https://github.com/numpy/numpy/pull/21442>`__: BUG: Ensure compile errors are raised correctly +* `#21450 <https://github.com/numpy/numpy/pull/21450>`__: PERF: Statically allocate unicode strings of memhandler +* `#21451 <https://github.com/numpy/numpy/pull/21451>`__: DOC: Style version switcher button +* `#21452 <https://github.com/numpy/numpy/pull/21452>`__: TST: Remove most prints from the test suit run +* `#21453 <https://github.com/numpy/numpy/pull/21453>`__: [road-to-cxx] npy_cpu_features moved to pure C +* `#21456 <https://github.com/numpy/numpy/pull/21456>`__: DOC: style main page card +* `#21463 <https://github.com/numpy/numpy/pull/21463>`__: BENCH: Add benchmarks targeted at small arrays +* `#21464 <https://github.com/numpy/numpy/pull/21464>`__: PERF: Fast check on equivalent arrays in PyArray_EQUIVALENTLY_ITERABLE_OVERLAP_OK +* `#21465 <https://github.com/numpy/numpy/pull/21465>`__: PERF: Use python integer on _count_reduce_items +* `#21466 <https://github.com/numpy/numpy/pull/21466>`__: DEV: Pin setuptools in the asv config +* `#21467 <https://github.com/numpy/numpy/pull/21467>`__: MAINT: Mark ``npy_memchr`` with ``no_sanitize("alignment")`` on clang +* `#21470 <https://github.com/numpy/numpy/pull/21470>`__: PERF: Skip probing ``__array_ufunc__`` for NumPy builtin scalars +* `#21477 <https://github.com/numpy/numpy/pull/21477>`__: MAINT: Reduce allocation size of empty (0 size) arrays to 1 byte +* `#21479 <https://github.com/numpy/numpy/pull/21479>`__: TYP,ENH: Add annotations for new numpy 1.23 features +* `#21485 <https://github.com/numpy/numpy/pull/21485>`__: ENH: Add 'keepdims' to 'average()' and 'ma.average()'. +* `#21490 <https://github.com/numpy/numpy/pull/21490>`__: TYP: Add typing for the keepdims param. of 'average' and 'ma.average' +* `#21491 <https://github.com/numpy/numpy/pull/21491>`__: DOC: Proposal - make the doc landing page cards more similar... +* `#21492 <https://github.com/numpy/numpy/pull/21492>`__: BUG: lib: Allow type uint64 for eye() arguments. +* `#21498 <https://github.com/numpy/numpy/pull/21498>`__: ENH: Add ``_get_madvise_hugepage`` function +* `#21499 <https://github.com/numpy/numpy/pull/21499>`__: ENH: avoid looping when dimensions[0] == 0 or array.size == 0 +* `#21500 <https://github.com/numpy/numpy/pull/21500>`__: TST: Fix uninitialized value in masked ndenumerate test +* `#21502 <https://github.com/numpy/numpy/pull/21502>`__: DEV: Fix Warnings/Errors on Gitpod +* `#21503 <https://github.com/numpy/numpy/pull/21503>`__: TYP: Add basic ``np.number`` overloads for ``ndarray`` dunders +* `#21514 <https://github.com/numpy/numpy/pull/21514>`__: MAINT: Update to Cython 0.29.29. +* `#21517 <https://github.com/numpy/numpy/pull/21517>`__: MAINT: Update .mailmap +* `#21518 <https://github.com/numpy/numpy/pull/21518>`__: BUG: Fix complex+longdouble and broken subclass handling +* `#21530 <https://github.com/numpy/numpy/pull/21530>`__: MAINT: Update to Cython 0.29.30. +* `#21534 <https://github.com/numpy/numpy/pull/21534>`__: BUG: Fix GCC error during build configuration +* `#21540 <https://github.com/numpy/numpy/pull/21540>`__: BUILD: update OpenBLAS to v0.3.20 +* `#21542 <https://github.com/numpy/numpy/pull/21542>`__: DOC: improve the docstring of numpy.sinc to explain behavior... +* `#21543 <https://github.com/numpy/numpy/pull/21543>`__: TST,TYP: Fix a python 3.11 failure for the ``GenericAlias`` tests +* `#21545 <https://github.com/numpy/numpy/pull/21545>`__: Tests/Docs: Update tests to Cython 0.29.30, mention in docs +* `#21552 <https://github.com/numpy/numpy/pull/21552>`__: BLD: Sort svml objects to keep builds reproducible +* `#21553 <https://github.com/numpy/numpy/pull/21553>`__: PERF: Faster MyPyFloat_AsDouble +* `#21558 <https://github.com/numpy/numpy/pull/21558>`__: MAINT: Python <3.8 related cleanups +* `#21562 <https://github.com/numpy/numpy/pull/21562>`__: REL: Update main after 1.22.4 release. +* `#21565 <https://github.com/numpy/numpy/pull/21565>`__: DOC: add explanation to makefile error +* `#21566 <https://github.com/numpy/numpy/pull/21566>`__: DOC: Fix docstring and examples for rfn.get_names* +* `#21568 <https://github.com/numpy/numpy/pull/21568>`__: DOC:linalg: Remove ref to scipy.linalg.pinv2 +* `#21569 <https://github.com/numpy/numpy/pull/21569>`__: MAINT: loosen Cython pin in environment.yml +* `#21570 <https://github.com/numpy/numpy/pull/21570>`__: CI: fix Gitpod image build +* `#21574 <https://github.com/numpy/numpy/pull/21574>`__: BUG: refguide-check: respect the verbosity +* `#21577 <https://github.com/numpy/numpy/pull/21577>`__: MAINT: update PyPy to 7.3.9 and remove unused script +* `#21580 <https://github.com/numpy/numpy/pull/21580>`__: MAINT: Update the cversion hash. +* `#21589 <https://github.com/numpy/numpy/pull/21589>`__: REL: Prepare for the NumPy 1.23.0rc1 release. +* `#21604 <https://github.com/numpy/numpy/pull/21604>`__: BUILD: fix tag name for travis: it is v1.23.0rc1 +* `#21606 <https://github.com/numpy/numpy/pull/21606>`__: DOC: add missing links for two NEPs +* `#21607 <https://github.com/numpy/numpy/pull/21607>`__: TYP, MAINT: Allow unsigned integer inplace-ops to accept signed... +* `#21610 <https://github.com/numpy/numpy/pull/21610>`__: REL: Prepare for 1.23.0rc1 release, second version. +* `#21619 <https://github.com/numpy/numpy/pull/21619>`__: MAINT, STY: Make download-wheels download source files. +* `#21634 <https://github.com/numpy/numpy/pull/21634>`__: MAINT: back out conversion of npymath component to c++ +* `#21635 <https://github.com/numpy/numpy/pull/21635>`__: TST: Skip F2PY tests without Fortran compilers +* `#21636 <https://github.com/numpy/numpy/pull/21636>`__: API: Retain ``arr.base`` more strictly in ``np.frombuffer`` +* `#21637 <https://github.com/numpy/numpy/pull/21637>`__: REL: Prepare for the NumPy 1.23.0rc2 release. +* `#21646 <https://github.com/numpy/numpy/pull/21646>`__: ENH: Add equals_nan kwarg to np.unique +* `#21649 <https://github.com/numpy/numpy/pull/21649>`__: MAINT: Start testing with Python 3.11. +* `#21656 <https://github.com/numpy/numpy/pull/21656>`__: TYP, ENH: Add annotations for the ``equal_nan`` keyword to ``np.unique`` +* `#21660 <https://github.com/numpy/numpy/pull/21660>`__: MAINT: Adapt npt._GenericAlias to Python 3.11 types.GenericAlias +* `#21684 <https://github.com/numpy/numpy/pull/21684>`__: MAINT: Point documentation version switcher at the docs homepage +* `#21688 <https://github.com/numpy/numpy/pull/21688>`__: DEP: Deprecate (rather than remove) the int-via-float parsing... +* `#21697 <https://github.com/numpy/numpy/pull/21697>`__: BUG: Fix a refactor leftover bug +* `#21698 <https://github.com/numpy/numpy/pull/21698>`__: BUG: Prevent attempted broadcasting of 0-D output operands in... +* `#21710 <https://github.com/numpy/numpy/pull/21710>`__: TST: Fixup loadtxt int-via-float tests when in release mode +* `#21716 <https://github.com/numpy/numpy/pull/21716>`__: ENH: Implement string comparison ufuncs (or almost) +* `#21718 <https://github.com/numpy/numpy/pull/21718>`__: BUG: use explicit einsum_path whenever it is given +* `#21719 <https://github.com/numpy/numpy/pull/21719>`__: BUG: Small fixupes found using valgrind +* `#21720 <https://github.com/numpy/numpy/pull/21720>`__: BUG: Enable fortran preprocessing for ifort on Windows +* `#21721 <https://github.com/numpy/numpy/pull/21721>`__: BLD, SIMD: Fix detect armhf and hardened the Neon/ASIMD compile-time... +* `#21722 <https://github.com/numpy/numpy/pull/21722>`__: BUG: .f2py_f2cmap doesn't map long_long and other options +* `#21729 <https://github.com/numpy/numpy/pull/21729>`__: REL: Prepare for the NumPy 1.23.0rc3 release. +* `#21754 <https://github.com/numpy/numpy/pull/21754>`__: BUG, SIMD: Fix detecting NEON/ASIMD on aarch64 +* `#21757 <https://github.com/numpy/numpy/pull/21757>`__: BUG: Do not skip value-based promotion path for large Python... +* `#21761 <https://github.com/numpy/numpy/pull/21761>`__: BUG: Fix small reference leaks found with pytest-leaks +* `#21777 <https://github.com/numpy/numpy/pull/21777>`__: REV: Revert "ENH: Implement string comparison ufuncs (or almost)... +* `#21809 <https://github.com/numpy/numpy/pull/21809>`__: MAINT: Add a check of the return value of PyMem_Calloc(). +* `#21810 <https://github.com/numpy/numpy/pull/21810>`__: BUG: lib: A loadtxt error message had two values reversed. +* `#21811 <https://github.com/numpy/numpy/pull/21811>`__: REL: Prepare for the NumPy 1.23.0 release +* `#21824 <https://github.com/numpy/numpy/pull/21824>`__: MAINT: Try fixing broken Anaconda uploads + + diff --git a/doc/source/release/1.23.0-notes.rst b/doc/source/release/1.23.0-notes.rst index 330e7fd44..c5c23620a 100644 --- a/doc/source/release/1.23.0-notes.rst +++ b/doc/source/release/1.23.0-notes.rst @@ -4,42 +4,409 @@ NumPy 1.23.0 Release Notes ========================== +The NumPy 1.23.0 release continues the ongoing work to improve the handling and +promotion of dtypes, increase the execution speed, clarify the documentation, +and expire old deprecations. The highlights are: -Highlights -========== +* Implementation of ``loadtxt`` in C, greatly improving its performance. +* Exposing DLPack at the Python level for easy data exchange. +* Changes to the promotion and comparisons of structured dtypes. +* Improvements to f2py. + +See below for the details, New functions ============= +* A masked array specialization of ``ndenumerate`` is now available as + ``numpy.ma.ndenumerate``. It provides an alternative to ``numpy.ndenumerate`` + and skips masked values by default. + + (`gh-20020 <https://github.com/numpy/numpy/pull/20020>`__) + +* ``numpy.from_dlpack`` has been added to allow easy exchange of data using the + DLPack protocol. It accepts Python objects that implement the ``__dlpack__`` + and ``__dlpack_device__`` methods and returns a ndarray object which is + generally the view of the data of the input object. + + (`gh-21145 <https://github.com/numpy/numpy/pull/21145>`__) + Deprecations ============ +* Setting ``__array_finalize__`` to ``None`` is deprecated. It must now be + a method and may wish to call ``super().__array_finalize__(obj)`` after + checking for ``None`` or if the NumPy version is sufficiently new. + + (`gh-20766 <https://github.com/numpy/numpy/pull/20766>`__) + +* Using ``axis=32`` (``axis=np.MAXDIMS``) in many cases had the + same meaning as ``axis=None``. This is deprecated and ``axis=None`` + must be used instead. + + (`gh-20920 <https://github.com/numpy/numpy/pull/20920>`__) + +* The hook function ``PyDataMem_SetEventHook`` has been deprecated and the + demonstration of its use in tool/allocation_tracking has been removed. The + ability to track allocations is now built-in to python via ``tracemalloc``. + + (`gh-20394 <https://github.com/numpy/numpy/pull/20394>`__) + +* ``numpy.distutils`` has been deprecated, as a result of ``distutils`` itself + being deprecated. It will not be present in NumPy for Python >= 3.12, and + will be removed completely 2 years after the release of Python 3.12 For more + details, see :ref:`distutils-status-migration`. + + (`gh-20875 <https://github.com/numpy/numpy/pull/20875>`__) -Future Changes -============== +* ``numpy.loadtxt`` will now give a ``DeprecationWarning`` when an integer + ``dtype`` is requested but the value is formatted as a floating point number. + + (`gh-21663 <https://github.com/numpy/numpy/pull/21663>`__) Expired deprecations ==================== +* The ``NpzFile.iteritems()`` and ``NpzFile.iterkeys()`` methods have been + removed as part of the continued removal of Python 2 compatibility. This + concludes the deprecation from 1.15. -Compatibility notes -=================== + (`gh-16830 <https://github.com/numpy/numpy/pull/16830>`__) +* The ``alen`` and ``asscalar`` functions have been removed. -C API changes -============= + (`gh-20414 <https://github.com/numpy/numpy/pull/20414>`__) + +* The ``UPDATEIFCOPY`` array flag has been removed together with the enum + ``NPY_ARRAY_UPDATEIFCOPY``. The associated (and deprecated) + ``PyArray_XDECREF_ERR`` was also removed. These were all deprecated in 1.14. They + are replaced by ``WRITEBACKIFCOPY``, that requires calling + ``PyArray_ResoveWritebackIfCopy`` before the array is deallocated. + + (`gh-20589 <https://github.com/numpy/numpy/pull/20589>`__) + +* Exceptions will be raised during array-like creation. When an object raised + an exception during access of the special attributes ``__array__`` or + ``__array_interface__``, this exception was usually ignored. This behaviour + was deprecated in 1.21, and the exception will now be raised. + + (`gh-20835 <https://github.com/numpy/numpy/pull/20835>`__) + +* Multidimensional indexing with non-tuple values is not allowed. Previously, + code such as ``arr[ind]`` where ``ind = [[0, 1], [0, 1]]`` produced a + ``FutureWarning`` and was interpreted as a multidimensional index (i.e., + ``arr[tuple(ind)]``). Now this example is treated like an array index over a + single dimension (``arr[array(ind)]``). Multidimensional indexing with + anything but a tuple was deprecated in NumPy 1.15. + + (`gh-21029 <https://github.com/numpy/numpy/pull/21029>`__) + +* Changing to a dtype of different size in F-contiguous arrays is no longer + permitted. Deprecated since Numpy 1.11.0. See below for an extended + explanation of the effects of this change. + + (`gh-20722 <https://github.com/numpy/numpy/pull/20722>`__) New Features ============ +crackfortran has support for operator and assignment overloading +---------------------------------------------------------------- +``crackfortran`` parser now understands operator and assignment +definitions in a module. They are added in the ``body`` list of the +module which contains a new key ``implementedby`` listing the names +of the subroutines or functions implementing the operator or +assignment. + +(`gh-15006 <https://github.com/numpy/numpy/pull/15006>`__) + +f2py supports reading access type attributes from derived type statements +------------------------------------------------------------------------- +As a result, one does not need to use ``public`` or ``private`` statements to +specify derived type access properties. + +(`gh-15844 <https://github.com/numpy/numpy/pull/15844>`__) + +New parameter ``ndmin`` added to ``genfromtxt`` +------------------------------------------------------------------------- +This parameter behaves the same as ``ndmin`` from ``numpy.loadtxt``. + +(`gh-20500 <https://github.com/numpy/numpy/pull/20500>`__) + +``np.loadtxt`` now supports quote character and single converter function +------------------------------------------------------------------------- +``numpy.loadtxt`` now supports an additional ``quotechar`` keyword argument +which is not set by default. Using ``quotechar='"'`` will read quoted fields +as used by the Excel CSV dialect. + +Further, it is now possible to pass a single callable rather than a dictionary +for the ``converters`` argument. + +(`gh-20580 <https://github.com/numpy/numpy/pull/20580>`__) + +Changing to dtype of a different size now requires contiguity of only the last axis +----------------------------------------------------------------------------------- +Previously, viewing an array with a dtype of a different item size required that +the entire array be C-contiguous. This limitation would unnecessarily force the +user to make contiguous copies of non-contiguous arrays before being able to +change the dtype. + +This change affects not only ``ndarray.view``, but other construction +mechanisms, including the discouraged direct assignment to ``ndarray.dtype``. + +This change expires the deprecation regarding the viewing of F-contiguous +arrays, described elsewhere in the release notes. + +(`gh-20722 <https://github.com/numpy/numpy/pull/20722>`__) + +Deterministic output files for F2PY +----------------------------------- +For F77 inputs, ``f2py`` will generate ``modname-f2pywrappers.f`` +unconditionally, though these may be empty. For free-form inputs, +``modname-f2pywrappers.f``, ``modname-f2pywrappers2.f90`` will both be generated +unconditionally, and may be empty. This allows writing generic output rules in +``cmake`` or ``meson`` and other build systems. Older behavior can be restored +by passing ``--skip-empty-wrappers`` to ``f2py``. :ref:`f2py-meson` details usage. + +(`gh-21187 <https://github.com/numpy/numpy/pull/21187>`__) + +``keepdims`` parameter for ``average`` +-------------------------------------- +The parameter ``keepdims`` was added to the functions ``numpy.average`` +and ``numpy.ma.average``. The parameter has the same meaning as it +does in reduction functions such as ``numpy.sum`` or ``numpy.mean``. + +(`gh-21485 <https://github.com/numpy/numpy/pull/21485>`__) + +New parameter ``equal_nan`` added to ``np.unique`` +-------------------------------------------------- +``np.unique`` was changed in 1.21 to treat all ``NaN`` values as equal and return +a single ``NaN``. Setting ``equal_nan=False`` will restore pre-1.21 behavior +to treat ``NaNs`` as unique. Defaults to ``True``. + +(`gh-21623 <https://github.com/numpy/numpy/pull/21623>`__) + + +Compatibility notes +=================== + +1D ``np.linalg.norm`` preserves float input types, even for scalar results +-------------------------------------------------------------------------- +Previously, this would promote to ``float64`` when the ``ord`` argument was +not one of the explicitly listed values, e.g. ``ord=3``:: + + >>> f32 = np.float32([1, 2]) + >>> np.linalg.norm(f32, 2).dtype + dtype('float32') + >>> np.linalg.norm(f32, 3) + dtype('float64') # numpy 1.22 + dtype('float32') # numpy 1.23 + +This change affects only ``float32`` and ``float16`` vectors with ``ord`` +other than ``-Inf``, ``0``, ``1``, ``2``, and ``Inf``. + +(`gh-17709 <https://github.com/numpy/numpy/pull/17709>`__) + +Changes to structured (void) dtype promotion and comparisons +------------------------------------------------------------ +In general, NumPy now defines correct, but slightly limited, promotion for +structured dtypes by promoting the subtypes of each field instead of raising +an exception:: + + >>> np.result_type(np.dtype("i,i"), np.dtype("i,d")) + dtype([('f0', '<i4'), ('f1', '<f8')]) + +For promotion matching field names, order, and titles are enforced, however +padding is ignored. +Promotion involving structured dtypes now always ensures native byte-order for +all fields (which may change the result of ``np.concatenate``) +and ensures that the result will be "packed", i.e. all fields are ordered +contiguously and padding is removed. +See :ref:`structured_dtype_comparison_and_promotion` for further details. + +The ``repr`` of aligned structures will now never print the long form including +``offsets`` and ``itemsize`` unless the structure includes padding not +guaranteed by ``align=True``. + +In alignment with the above changes to the promotion logic, the +casting safety has been updated: + +* ``"equiv"`` enforces matching names and titles. The itemsize + is allowed to differ due to padding. +* ``"safe"`` allows mismatching field names and titles +* The cast safety is limited by the cast safety of each included + field. +* The order of fields is used to decide cast safety of each + individual field. Previously, the field names were used and + only unsafe casts were possible when names mismatched. + +The main important change here is that name mismatches are now +considered "safe" casts. + +(`gh-19226 <https://github.com/numpy/numpy/pull/19226>`__) + +``NPY_RELAXED_STRIDES_CHECKING`` has been removed +------------------------------------------------- +NumPy cannot be compiled with ``NPY_RELAXED_STRIDES_CHECKING=0`` +anymore. Relaxed strides have been the default for many years and +the option was initially introduced to allow a smoother transition. + +(`gh-20220 <https://github.com/numpy/numpy/pull/20220>`__) + +``np.loadtxt`` has recieved several changes +------------------------------------------- + +The row counting of ``numpy.loadtxt`` was fixed. ``loadtxt`` ignores fully +empty lines in the file, but counted them towards ``max_rows``. +When ``max_rows`` is used and the file contains empty lines, these will now +not be counted. Previously, it was possible that the result contained fewer +than ``max_rows`` rows even though more data was available to be read. +If the old behaviour is required, ``itertools.islice`` may be used:: + + import itertools + lines = itertools.islice(open("file"), 0, max_rows) + result = np.loadtxt(lines, ...) + +While generally much faster and improved, ``numpy.loadtxt`` may now fail to +converter certain strings to numbers that were previously successfully read. +The most important cases for this are: + +* Parsing floating point values such as ``1.0`` into integers is now deprecated. +* Parsing hexadecimal floats such as ``0x3p3`` will fail +* An ``_`` was previously accepted as a thousands delimiter ``100_000``. + This will now result in an error. + +If you experience these limitations, they can all be worked around by passing +appropriate ``converters=``. NumPy now supports passing a single converter +to be used for all columns to make this more convenient. +For example, ``converters=float.fromhex`` can read hexadecimal float numbers +and ``converters=int`` will be able to read ``100_000``. + +Further, the error messages have been generally improved. However, this means +that error types may differ. In particularly, a ``ValueError`` is now always +raised when parsing of a single entry fails. + +(`gh-20580 <https://github.com/numpy/numpy/pull/20580>`__) + Improvements ============ +``ndarray.__array_finalize__`` is now callable +---------------------------------------------- +This means subclasses can now use ``super().__array_finalize__(obj)`` +without worrying whether ``ndarray`` is their superclass or not. +The actual call remains a no-op. + +(`gh-20766 <https://github.com/numpy/numpy/pull/20766>`__) + +Add support for VSX4/Power10 +---------------------------------------------- +With VSX4/Power10 enablement, the new instructions available in +Power ISA 3.1 can be used to accelerate some NumPy operations, +e.g., floor_divide, modulo, etc. + +(`gh-20821 <https://github.com/numpy/numpy/pull/20821>`__) + +``np.fromiter`` now accepts objects and subarrays +------------------------------------------------- +The ``numpy.fromiter`` function now supports object and +subarray dtypes. Please see he function documentation for +examples. + +(`gh-20993 <https://github.com/numpy/numpy/pull/20993>`__) + +Math C library feature detection now uses correct signatures +------------------------------------------------------------ +Compiling is preceded by a detection phase to determine whether the +underlying libc supports certain math operations. Previously this code +did not respect the proper signatures. Fixing this enables compilation +for the ``wasm-ld`` backend (compilation for web assembly) and reduces +the number of warnings. + +(`gh-21154 <https://github.com/numpy/numpy/pull/21154>`__) + +``np.kron`` now maintains subclass information +---------------------------------------------- +``np.kron`` maintains subclass information now such as masked arrays +while computing the Kronecker product of the inputs + +.. code-block:: python + + >>> x = ma.array([[1, 2], [3, 4]], mask=[[0, 1], [1, 0]]) + >>> np.kron(x,x) + masked_array( + data=[[1, --, --, --], + [--, 4, --, --], + [--, --, 4, --], + [--, --, --, 16]], + mask=[[False, True, True, True], + [ True, False, True, True], + [ True, True, False, True], + [ True, True, True, False]], + fill_value=999999) + +.. warning:: + ``np.kron`` output now follows ``ufunc`` ordering (``multiply``) + to determine the output class type + + .. code-block:: python + + >>> class myarr(np.ndarray): + >>> __array_priority__ = -1 + >>> a = np.ones([2, 2]) + >>> ma = myarray(a.shape, a.dtype, a.data) + >>> type(np.kron(a, ma)) == np.ndarray + False # Before it was True + >>> type(np.kron(a, ma)) == myarr + True + +(`gh-21262 <https://github.com/numpy/numpy/pull/21262>`__) + + +Performance improvements and changes +==================================== + +Faster ``np.loadtxt`` +--------------------- +``numpy.loadtxt`` is now generally much faster than previously as most of it +is now implemented in C. + +(`gh-20580 <https://github.com/numpy/numpy/pull/20580>`__) + +Faster reduction operators +-------------------------- +Reduction operations like ``numpy.sum``, ``numpy.prod``, ``numpy.add.reduce``, +``numpy.logical_and.reduce`` on contiguous integer-based arrays are now +much faster. + +(`gh-21001 <https://github.com/numpy/numpy/pull/21001>`__) + +Faster ``np.where`` +------------------- +``numpy.where`` is now much faster than previously on unpredictable/random +input data. + +(`gh-21130 <https://github.com/numpy/numpy/pull/21130>`__) + +Faster operations on NumPy scalars +---------------------------------- +Many operations on NumPy scalars are now significantly faster, although +rare operations (e.g. with 0-D arrays rather than scalars) may be slower +in some cases. +However, even with these improvements users who want the best performance +for their scalars, may want to convert a known NumPy scalar into a Python +one using ``scalar.item()``. + +(`gh-21188 <https://github.com/numpy/numpy/pull/21188>`__) + +Faster ``np.kron`` +------------------ +``numpy.kron`` is about 80% faster as the product is now computed +using broadcasting. -Changes -======= +(`gh-21354 <https://github.com/numpy/numpy/pull/21354>`__) |