1.16.0-changelog.rst

Contributors

A total of 113 people contributed to this release. People with a "+" by their names contributed a patch for the first time.

• Alan Fontenot +
• Allan Haldane
• Alon Hershenhorn +
• Alyssa Quek +
• Andreas Nussbaumer +
• Anner +
• Anthony Sottile +
• Antony Lee
• Ayappan P +
• Bas van Schaik +
• C.A.M. Gerlach +
• Charles Harris
• Chris Billington
• Christian Clauss
• Christoph Gohlke
• Christopher Pezley +
• Daniel B Allan +
• Daniel Smith
• Dawid Zych +
• Derek Kim +
• Dima Pasechnik +
• Edgar Giovanni Lepe +
• Elena Mokeeva +
• Elliott Sales de Andrade +
• Emil Hessman +
• Eric Larson
• Eric Schles +
• Eric Wieser
• Giulio Benetti +
• Guillaume Gautier +
• Guo Ci
• Heath Henley +
• Isuru Fernando +
• 10. Lewis Muir +
• Jack Vreeken +
• Jaime Fernandez
• James Bourbeau
• Jeff VanOss
• Jeffrey Yancey +
• Jeremy Chen +
• Jeremy Manning +
• Jeroen Demeyer
• John Darbyshire +
• John Kirkham
• John Zwinck
• Jonas Jensen +
• Joscha Reimer +
• Juan Azcarreta +
• Julian Taylor
• Kevin Sheppard
• Krzysztof Chomski +
• Kyle Sunden
• Lars Grüter
• Lilian Besson +
• MSeifert04
• Mark Harfouche
• Marten van Kerkwijk
• Martin Thoma
• Matt Harrigan +
• Matthew Bowden +
• Matthew Brett
• Matthias Bussonnier
• Matti Picus
• Max Aifer +
• Michael Hirsch, Ph.D +
• Michael James Jamie Schnaitter +
• MichaelSaah +
• Mike Toews
• Minkyu Lee +
• Mircea Akos Bruma +
• Mircea-Akos Brumă +
• Moshe Looks +
• Muhammad Kasim +
• Nathaniel J. Smith
• Nikita Titov +
• Paul Müller +
• Paul van Mulbregt
• Pauli Virtanen
• Pierre Glaser +
• Pim de Haan
• Ralf Gommers
• Robert Kern
• Robin Aggleton +
• Rohit Pandey +
• Roman Yurchak +
• Ryan Soklaski
• Sebastian Berg
• Sho Nakamura +
• Simon Gibbons
• Stan Seibert +
• Stefan Otte
• Stefan van der Walt
• Stephan Hoyer
• Stuart Archibald
• Taylor Smith +
• Tim Felgentreff +
• Tim Swast +
• Tim Teichmann +
• Toshiki Kataoka
• Travis Oliphant
• Tyler Reddy
• Uddeshya Singh +
• Warren Weckesser
• Weitang Li +
• Wenjamin Petrenko +
• William D. Irons
• Yannick Jadoul +
• Yaroslav Halchenko
• Yug Khanna +
• Yuji Kanagawa +
• Yukun Guo +
• @ankokumoyashi +
• @lerbuke +

Pull requests merged

A total of 490 pull requests were merged for this release.

• #6256: NEP: Add proposal for oindex and vindex.
• #6377: BUG: define "uint-alignment", fixes complex64 alignment
• #8206: ENH: add padding options to diff
• #8923: ENH: Add 'stone' estimator to np.histogram
• #8955: ENH: Allow ufunc.identity to be any python object
• #9022: BUG: don't silence __array_wrap__ errors in ufunc.reduce
• #10551: BUG: memmap close files when it shouldn't, load leaves them open...
• #10602: MAINT: Move dtype string functions to python
• #10704: NEP 15: Merging multiarray and umath
• #10797: DEP: Updated unravel_index() to support shape kwarg
• #10915: ENH: implement nep 0015: merge multiarray and umath
• #10998: DOC: removed spurious FIXME comment in number.c
• #11002: MAINT: add clearer message to assist users with failed builds.
• #11016: ENH: Add AARCH32 support.
• #11084: DOC: link to TESTS.rst.txt testing guidelines, tweak testing...
• #11119: ENH: Chain exceptions to give better error messages for invalid...
• #11175: ENH: Generalized ufunc signature expansion for frozen and flexible...
• #11197: BUG/ENH: Removed non-standard scaling of the covariance matrix...
• #11234: DOC: Update einsum docs
• #11282: MAINT: move comparison operator special-handling out of ufunc...
• #11297: NEP: Expansion of gufunc signatures.
• #11299: BUG: Prevent crashes on 0-length structured void scalars
• #11303: DOC: revision of NEP-18 (__array_function__)
• #11312: WIP: DOC: slightly tweak the directions to create a release
• #11318: REL: Setup master for 1.16 development.
• #11323: DEP: Actually deprecate the normed argument to histogram
• #11324: MAINT: Don't use dtype strings when the dtypes themselves can...
• #11326: DOC: Update master after NumPy 1.14.5 release.
• #11328: MAINT: Misc numeric cleanup
• #11335: DOC: Change array lengths/entries in broadcast_arrays example...
• #11336: BUG: decref in failure path; replace PyObject_Type by Py_TYPE
• #11338: MAINT: Ensure ufunc override call each class only once, plus...
• #11340: BUG: sctypeDict['f8'] randomly points to double or longdouble...
• #11345: BUG/ENH: Einsum optimization path updates and bug fixes.
• #11347: DOC: Silence many sphinx warnings
• #11348: ENH: Improve support for pathlib.Path objects in load functions
• #11349: DOC: document new functions
• #11351: MAINT: Improve speed of ufunc kwargs parsing
• #11353: DOC, MAINT: HTTP -> HTTPS, and other linkrot fixes
• #11356: NEP: Update NEP 19: RNG Policy
• #11357: MAINT: Add new _test.c files and benchmarks/html to gitignore
• #11365: BUG: add missing NpyIter_Close in einsum
• #11366: BUG/TST: String indexing should just fail, not emit a futurewarning
• #11371: DOC: Clarify requirement that histogram bins are monotonic.
• #11373: TST: Show that histogramdd's normed argument is histogram's density
• #11374: WIP: additional revision for NEP-18 (__array_function__)
• #11376: ENH: Remove NpyIter_Close
• #11379: BUG: changed hardcoded axis to 0 for checking indices
• #11382: DEP: deprecate undocumented, unused dtype type dicts
• #11383: ENH: Allow size=0 in numpy.random.choice
• #11385: BUG: Make scalar.squeeze accept axis arg
• #11390: REL,MAINT: Update numpyconfig.h for 1.15.
• #11391: MAINT: Update mailmap
• #11396: TST: Added regression test for #11395
• #11405: BUG: Ensure comparisons on scalar strings pass without warning.
• #11406: BUG: Ensure out is returned in einsum.
• #11409: DOC: Update testing section of README.
• #11414: DOC: major revision of NEP 21, advanced indexing
• #11422: BENCH: Add benchmarks for np.loadtxt reading from CSV format...
• #11424: ENH: Allow use of svd on empty arrays
• #11425: DOC: Clear up confusion between np.where(cond) and np.where(cond,...
• #11428: BUG: Fix incorrect deprecation logic for histogram(normed=...)...
• #11429: NEP: accept NEP 20 partially (frozen, flexible, but not broadcastable...
• #11432: MAINT: Refactor differences between cblas_matrixproduct and PyArray_MatrixProduct2
• #11434: MAINT: add PyPI classifier for Python 3.7
• #11436: DOC: Document average return type
• #11440: BUG: fix interpolation with inf and NaN present
• #11444: DOC: Fix documentation for fromfunction
• #11449: BUG: Revert #10229 to fix DLL loads on Windows.
• #11450: MAINT/DEP: properly implement ndarray.__pos__
• #11453: BENCH: add ufunc argument parsing benchmarks.
• #11455: BENCH: belated addition of lcm, gcd to ufunc benchmark.
• #11459: NEP: Add some text to NEP 0 to clarify how a NEP is accepted
• #11461: MAINT: Add discussion link to NEP 15
• #11462: Add NEP 22, a high level overview for the duck array work
• #11463: MAINT: Produce a more readable repr of argument packs in benchmark
• #11464: BUG: Don't convert inputs to np.float64 in digitize
• #11468: BUG: Advanced indexing assignment incorrectly took 1-D fastpath
• #11470: BLD: Don't leave the build task running if runtests.py is interrupted
• #11471: MAINT: Remove python-side docstrings from add_newdocs.
• #11472: DOC: include NEP number on each NEP page
• #11473: MAINT: Move pytesttester outside of np.testing, to avoid creating...
• #11474: MAINT: Move add_newdocs into core, since it only adds docs to...
• #11479: BUG: Fix #define for ppc64 and ppc64le
• #11480: MAINT: move ufunc override code to umath and multiarray as much...
• #11482: DOC: Include warning in np.resize() docs
• #11484: BUG: Increase required cython version on python 3.7
• #11487: DOC: extend sanity check message
• #11488: NEP: clarify bugfix policy for legacy RandomState.
• #11501: MAINT: Tidy cython invocation
• #11503: MAINT: improve error message for isposinf and isneginf on complex...
• #11512: DOC: Add templates for issues and PRs
• #11514: Prefer the same-python cython to the on-PATH cython
• #11515: BUG: decref of field title caused segfault
• #11518: MAINT: Speed up normalize_axis_tuple by about 30%
• #11522: BUG: fix np.load() of empty .npz file
• #11525: MAINT: Append *FLAGS instead of overriding
• #11526: ENH: add multi-field assignment helpers in np.lib.recfunctions
• #11527: DOC: Note that method is the polar form of Box-Muller.
• #11528: ENH: Add support for ipython latex printing to polynomial
• #11531: ENH: Add density argument to histogramdd.
• #11533: DOC: Fixed example code for cheb2poly and poly2cheb (see #11519)
• #11534: DOC: Minor improvements to np.concatenate docstring
• #11535: MAINT: Improve memory usage in PEP3118 format parsing
• #11553: DOC: Tiny typo on numpy/reference/arrays.dtypes.html
• #11556: BUG: Make assert_string_equal check str equality simply without...
• #11559: NEP: accept nep 0015
• #11560: NEP: accept nep 0019
• #11562: DOC: update release notes for LDFLAGS append behavior (gh-11525).
• #11565: MAINT: convert the doctests for polynomial to regular tests
• #11566: BLD: Do not use gcc warnings flags when 'gcc' is actually clang.
• #11567: TST: Integrate codecov testing
• #11568: BLD: Modify cpu detection and printing to get working aarch64...
• #11571: DOC: Updated array2string description
• #11572: DOC: Updated Slice Description
• #11573: TST: add broadcast_arrays() kwarg unit test for TypeError
• #11580: MAINT: refactor ufunc iter operand flags handling
• #11591: MAINT: update runtests.py node id example for pytest usage
• #11592: DOC: add Stefan van der Walt to Steering Council
• #11593: ENH: handle empty matrices in qr decomposition
• #11594: ENH: support for empty matrices in linalg.lstsq
• #11595: BUG:warn on Nan in minimum,maximum for scalars, float16
• #11596: NEP: backwards compatibility and deprecation policy
• #11598: TST: Add Python 3.7 to CI testing
• #11601: BUG: Make np.array([[1], 2]) and np.array([1, [2]]) behave in...
• #11606: DOC: Post 1.15.0 release updates for master.
• #11607: DOC: minor clarification and typo fix to NEP 21 (outer indexing).
• #11610: TST: including C source line coverage for CI / codecov
• #11611: NEP: Add roadmap section and subdocuments to NEPs
• #11613: BUG: have geometric() raise ValueError on p=0
• #11615: BUG: Clip uses wrong memory order in output
• #11616: DOC: add a brief note on "Protocols for methods" to NEP 18
• #11621: DOC: Use "real symmetric" rather than "symmetric" in eigh...
• #11626: DOC: Show plot in meshgrid example.
• #11630: DOC: Include the versionadded to the isnat documentation.
• #11634: MAINT: Filter Cython warnings in __init__.py
• #11637: ENH: np.angle: Remove unnecessary multiplication, and allow subclasses...
• #11638: ENH: Make expand_dims work on subclasses
• #11642: BUG: Fixes for unicode field names in Python 2
• #11643: DOC: Insert up to date link to Spyder website in Dev Env doc...
• #11644: BUG: Fix doc source links to unwrap decorators
• #11652: BUG: Ensure singleton dimensions are not dropped when converting...
• #11660: ENH: Add Nan warnings for maximum, minimum on more dtypes
• #11669: BUG: Fix regression in void_getitem
• #11670: MAINT: trivially refactor mapped indexing
• #11673: DOC: Add geomspace to "See also" of linspace
• #11679: TST: ignore setup.py files for codecov reports
• #11688: DOC: Update broadcasting doc with current exception details
• #11691: BUG: Make matrix_power again work for object arrays.
• #11692: MAINT: Remove duplicate code.
• #11693: NEP: Mark NEP 18 as accepted
• #11694: BUG: Fix pickle and memoryview for datetime64, timedelta64 scalars
• #11695: BUG: Add missing PyErr_NoMemory after failing malloc
• #11703: MAINT: Remove np.pkgload, which seems to be unusable anyway
• #11708: BUG: Fix regression in np.loadtxt for bz2 text files in Python...
• #11710: BUG: Check for compiler used in env['CC'], then config_vars['CC']
• #11711: BUG: Fix undefined functions on big-endian systems.
• #11715: TST: Fix urlopen stubbing.
• #11717: MAINT: Make einsum optimize default to False.
• #11718: BUG: Revert use of console_scripts.
• #11722: MAINT: Remove duplicate docstring and correct location of __all__...
• #11725: BUG: Fix Fortran kind detection for aarch64 & s390x.
• #11727: BUG: Fix printing of longdouble on ppc64le.
• #11729: DOC: fix capitalization of kilojoules
• #11731: DOC: fix typo in vectorize docstring
• #11733: DOC: recommend polynomial.Polynomial over np.polyfit
• #11735: BUG: Fix test sensitive to platform byte order.
• #11738: TST, MAINT: add lgtm.yml to tweak LGTM.com analysis
• #11739: BUG: disallow setting flag to writeable after fromstring, frombuffer
• #11740: BUG: Deprecation triggers segfault
• #11742: DOC: Reduce warnings and cleanup redundant c-api documentation
• #11745: DOC: Small docstring fixes for old polyfit.
• #11754: BUG: check return value of _buffer_format_string
• #11755: MAINT: Fix typos in random.hypergeometric's notes
• #11756: MAINT: Make assert_array_compare more generic.
• #11765: DOC: Move documentation from help(ndarray.ctypes) to help(some_array.ctypes)
• #11771: BUG: Make random.shuffle work on 1-D instances of ndarray...
• #11774: BUG: Fix regression in intersect1d.
• #11778: BUG: Avoid signed overflow in histogram
• #11783: MAINT: check _append_char return value
• #11784: MAINT: reformat line spacing before test methods
• #11797: DOC: Update docs after 1.15.1 release.
• #11800: DOC: document use when f2py is not in the PATH
• #11802: ENH: Use entry_points to install the f2py scripts.
• #11805: BUG: add type cast check for ediff1d
• #11806: DOC: Polybase augmented assignment notes
• #11812: DOC: edit setup.py docstring that is displayed on PyPI.
• #11813: BUG: fix array_split incorrect behavior with array size bigger...
• #11814: DOC, MAINT: Fixes for errstate() and README.md documentation.
• #11817: DOC: add examples and extend existing dos for polynomial subclasses
• #11818: TST: add missing tests for all polynomial subclass pow fns.
• #11823: TST: add test for array2string unexpected kwarg
• #11830: MAINT: reduce void type repr code duplication
• #11834: MAINT, DOC: Replace 'an' by 'a' in some docstrings.
• #11837: DOC: Make clear the connection between numpy types and C types
• #11840: BUG: Let 0-D arrays of Python timedelta convert to np.timedelta64.
• #11843: MAINT: remove surviving, unused, list comprehension
• #11849: TST: reorder duplicate mem_overlap.c compile
• #11850: DOC: add comment to remove fn after python 2 support is dropped
• #11852: BUG: timedelta64 now accepts NumPy ints
• #11858: DOC: add docstrings for numeric types
• #11862: BUG: Re-add _ones_like to numpy.core.umath.
• #11864: TST: Update travis testing to use latest virtualenv.
• #11865: DOC: add a Code of Conduct document.
• #11866: TST: Drop Python 3.4 testing
• #11868: MAINT: include benchmarks, complete docs, dev tool files in sdist.
• #11870: MAINT: dtype(unicode) should raise TypeError on failure
• #11874: BENCH: split out slow setup method in bench_shape_base.Block
• #11877: BUG: Fix memory leak in pyfragments.swg
• #11880: BUG: The multiarray/ufunc merge broke old wheels.
• #11882: DOC: Recommend the use of np.ndim over np.isscalar, and explain...
• #11889: BENCH: Split bench_function_base.Sort into Sort and SortWorst.
• #11891: MAINT: remove exec_command() from build_ext
• #11892: TST: Parametrize PEP3118 scalar tests.
• #11893: TST: Fix duplicated test name.
• #11894: TST: Parametrize f2py tests.
• #11895: TST: Parametrize some linalg tests over types.
• #11896: BUG: Fix matrix PendingDeprecationWarning suppression for pytest...
• #11898: MAINT: remove exec_command usage from ccompiler.py
• #11899: MAINT: remove exec_command from system_info.py
• #11900: MAINT: remove exec_command from gnu.py
• #11901: MAINT: remove exec_command usage in ibm.py
• #11904: Use pytest for some already-parametrized core tests
• #11905: TST: Start testing with "-std=c99" on travisCI.
• #11906: TST: add shippable ARMv8 to CI
• #11907: Link HOWTO_DOCUMENT to specific section on docstrings
• #11909: MAINT: flake8 cleanups
• #11910: MAINT: test, refactor design of recursive closures
• #11912: DOC: dtype offset and itemsize is limited by range of C int
• #11914: DOC: Clarify difference between PySequence_GETITEM, PyArray_GETITEM
• #11916: DEP: deprecate np.set_numeric_ops and friends
• #11920: TST: Fix 'def' test_numerictypes.py::TestSctypeDict to 'class'...
• #11921: MAINT: Don't rely on __name__ in bitname - use the information...
• #11922: TST: Add tests for maximum_sctype
• #11929: DOC: #defining -> #define / Added a short explanation for Numeric
• #11930: DOC: fix scipy-sphinx-theme license path
• #11932: MAINT: Move np.dtype.name.__get__ to python
• #11933: TST: Fix unit tests that used to call unittest.TestCase.fail
• #11934: NEP: Revert "NEP: Mark NEP 18 as accepted"
• #11935: MAINT: remove usage of exec_command in config.py
• #11937: MAINT: remove exec_command() from f2py init
• #11941: BUG: Ensure einsum(optimize=True) dispatches tensordot deterministically
• #11943: DOC: Add warning/clarification about backwards compat in NEP-18
• #11948: DEP: finish making all comparisons to NaT false
• #11949: MAINT: Small tidy-ups to np.core._dtype
• #11950: MAINT: Extract tangential improvements made in #11175
• #11952: MAINT: test NPY_INTERNAL_BUILD only if defined
• #11953: TST: codecov.yml improvements
• #11957: ENH: mark that large allocations can use huge pages
• #11958: TST: Add a test for np.pad where constant_values is an object
• #11959: MAINT: Explicitely cause pagefaults to happen before starting...
• #11961: TST: Add more tests for np.pad
• #11962: ENH: maximum lines of content to be read from numpy.loadtxt
• #11965: BENCH: Add a benchmark comparing block to copy in the 3D case
• #11966: MAINT: Rewrite shape normalization in pad function
• #11967: BUG: fix refcount leak in PyArray_AdaptFlexibleDType
• #11971: MAINT: Block algorithm with a single copy per call to block
• #11973: BUG: fix cached allocations without the GIL
• #11976: MAINT/DOC: Show the location of an empty list in np.block
• #11979: MAINT: Ensure that a copy of the array is returned when calling...
• #11989: BUG: Ensure boolean indexing of subclasses sets base correctly.
• #11991: MAINT: speed up _block by avoiding a recursive closure
• #11996: TST: Parametrize and break apart dtype tests
• #11997: MAINT: Extract string helpers to a new private file
• #12002: Revert "NEP: Revert "NEP: Mark NEP 18 as accepted""
• #12004: BUG: Fix f2py compile function testing.
• #12005: ENH: initial implementation of core __array_function__ machinery
• #12008: MAINT: Reassociate np.cast with the comment describing it
• #12009: MAINT: Eliminate the private numerictypes._typestr
• #12011: ENH: implementation of array_reduce_ex
• #12012: MAINT: Extract the crazy number of type aliases to their own...
• #12014: TST: prefer pytest.skip() over SkipTest
• #12015: TST: improve warnings parallel test safety
• #12017: NEP: add 3 missing data NEPs rescued from 2011-2012
• #12018: MAINT: Simplify parts of _type_aliases
• #12019: DOC: MAINT: address comments @eric-wieser on NEP 24-26 PR.
• #12020: TST: Add tests for np.sctype2char
• #12021: DOC: Post NumPy 1.15.2 release updates.[ci skip]
• #12024: MAINT: Normalize axes the normal way in fftpack.py
• #12027: DOC: Add docstrings for abstract types in scalar type hierarchy
• #12030: DOC: use "import numpy as np" style
• #12032: BUG: check return value from PyArray_PromoteTypes
• #12033: TST: Mark check for f2py script xfail.
• #12034: MAINT: Add version deprecated to some deprecation messages.
• #12035: BUG: Fix memory leak in PY3K buffer code.
• #12041: MAINT: remove duplicate imports
• #12042: MAINT: cleanup and better document core/overrides.py
• #12045: BUG: fix memory leak of buffer format string
• #12048: BLD: pin sphinx to 1.7.9
• #12051: TST: add macos azure testing to CI
• #12054: MAINT: avoid modifying mutable default values
• #12056: MAINT: The crackfortran function is called with an extra argument
• #12057: MAINT: remove unused imports
• #12058: MAINT: remove redundant assignment
• #12060: MAINT: remove unused stdlib imports
• #12061: MAINT: remove redundant imports
• #12062: BUG: OBJECT_to_* should check for errors
• #12064: MAINT: delay initialization of getlimits (circular imports)
• #12072: BUG: test_path() now uses Path.resolve()
• #12073: MAINT Avoid some memory copies in numpy.polynomial.hermite
• #12079: MAINT: Blacklist some MSVC complex functions.
• #12081: TST: add Windows test matrix to Azure CI
• #12082: TST: Add Python 3.5 to Azure windows CI.
• #12088: BUG: limit default for get_num_build_jobs() to 8
• #12089: BUG: Fix in-place permutation
• #12090: TST, MAINT: Update pickling tests by making them loop over all...
• #12091: TST: Install pickle5 for CI testing with python 3.6/7
• #12093: Provide information about what kind is actually not integer kind
• #12099: ENH: Validate dispatcher functions in array_function_dispatch
• #12102: TST: improve coverage of nd_grid
• #12103: MAINT: Add azure-pipeline status badge to README.md
• #12106: TST, MAINT: Skip some f2py tests on Mac.
• #12108: BUG: Allow boolean subtract in histogram
• #12109: TST: add unit test for issctype
• #12112: ENH: check getfield arguments to prevent invalid memory access
• #12115: ENH: __array_function__ support for most of numpy.core
• #12116: ENH: __array_function__ support for np.lib, part 1/2
• #12117: ENH: __array_function__ support for np.fft and np.linalg
• #12119: ENH: __array_function__ support for np.lib, part 2/2
• #12120: ENH: add timedelta modulus operator support (mm)
• #12121: MAINT: Clarify the error message for resize failure
• #12123: DEP: deprecate asscalar
• #12124: BUG: refactor float error status to support Alpine linux
• #12125: TST: expand cases in test_issctype()
• #12127: BUG: Fix memory leak in mapping.c
• #12131: BUG: fix PyDataType_ISBOOL
• #12133: MAINT, TST refactor pickle imports and tests
• #12134: DOC: Remove duplicated sentence in numpy.multiply
• #12137: TST: error tests for fill_diagonal()
• #12138: TST: error tests for diag_indices_from()
• #12140: DOC: fixups for NEP-18 based on the implementation
• #12141: DOC: minor tweak to CoC (update NumFOCUS contact address).
• #12145: MAINT: Update ndarrayobject.h __cplusplus block.
• #12146: MAINT: Fix typo in comment
• #12147: MAINT: Move duplicated type_reso_error code into a helper function
• #12148: DOC: document NEP-18 overrides in release notes
• #12151: TST: byte_bounds contiguity handling
• #12153: DOC, TST: cover setdiff1d assume_unique
• #12154: ENH: __array_function__ for np.core.defchararray
• #12155: MAINT: Define Py_SETREF for pre-3.5.2 python and use in code
• #12157: ENH: Add support for third-party path-like objects by backporting...
• #12159: MAINT: remove unused nd_grid __len__.
• #12163: ENH: __array_function__ for np.einsum and np.block
• #12165: Mark NEP 22 as accepted, and add "Informational" NEPs to NEP...
• #12166: NEP: Add zero-rank arrays historical info NEP
• #12173: NEP: add notes about updates to NEP-18
• #12174: NEP 16 abstract arrays: rebased and marked as "Withdrawn"
• #12175: ENH: __array_function__ for multiarray functions
• #12176: TST: add test for weighted histogram mismatch
• #12177: MAINT: remove unused _assertSquareness()
• #12179: MAINT: Move _kind_to_stem to np.core._dtype, so that it can...
• #12180: NEP: change toc titles, cross reference, mark 16 superseded
• #12181: MAINT: fix depreciation message typo for np.sum
• #12185: TST: test multi_dot with 2 arrays
• #12199: TST: add Azure CI triggers
• #12209: Delay import of distutils.msvccompiler to avoid warning on non-Windows.
• #12211: DOC: Clarify the examples for argmax and argmin
• #12212: MAINT: ndarray.__repr__ should not rely on __array_function__
• #12214: TST: add test for tensorinv()
• #12215: TST: test dims match on lstsq()
• #12216: TST: test invalid histogram range
• #12217: TST: test histogram bins dims
• #12219: ENH: make matmul into a ufunc
• #12222: TST: unit tests for column_stack.
• #12224: BUG: Fix MaskedArray fill_value type conversion.
• #12229: MAINT: Fix typo in comment
• #12236: BUG: maximum, minimum no longer emit warnings on NAN
• #12240: BUG: Fix crash in repr of void subclasses
• #12241: TST: arg handling tests in histogramdd
• #12243: BUG: Fix misleading assert message in assert_almost_equal #12200
• #12245: TST: tests for sort_complex()
• #12246: DOC: Update docs after NumPy 1.15.3 release.
• #12249: BUG: Dealloc cached buffer info
• #12250: DOC: add missing docs
• #12251: MAINT: improved error message when no __array_function__ implementation...
• #12254: MAINT: Move ctype -> dtype conversion to python
• #12257: BUG: Fix fill value in masked array '==' and '!=' ops.
• #12259: TST: simplify how the different code paths for block are tested.
• #12265: BUG: Revert linspace import for concatenation funcs
• #12266: BUG: Avoid SystemErrors by checking the return value of PyPrint
• #12268: DOC: add broadcasting article from scipy old-wiki
• #12270: MAINT: set __module__ for more array_function_dispatch uses
• #12276: MAINT: remove unused parse_index()
• #12279: NEP: tweak and mark NEP 0027 as final
• #12280: DEP: deprecate passing a generator to stack functions
• #12281: NEP: revise note for NEP 27
• #12285: ENH: array does not need to be writable to use as input to take
• #12286: ENH: Do not emit compiler warning if forcing old API
• #12288: BUILD: force LGTM to use cython>=0.29
• #12291: MAINT: _set_out_array() syntax fix
• #12292: MAINT: removed unused vars in f2py test code
• #12299: BUILD: use system python3 in the chroot
• #12302: DOC: Update the docstring of asfortranarray and ascontiguousarray
• #12306: TST: add 32-bit linux Azure CI job
• #12312: MAINT, TST: unreachable Python code paths
• #12321: MAINT: Simple speed-ups for getting overloaded types
• #12326: DOC: NumPy 1.15.4 post release documentation update.
• #12328: MAINT: Allow subclasses in ndarray.__array_function__.
• #12330: TST: test_tofile_fromfile now uses initialized memory
• #12331: DEV: change ASV benchmarks to run on Python 3.6 by default
• #12338: DOC: add a docstring for the function 'compare_chararrays' (See...
• #12342: BUG: Fix for np.dtype(ctypes.Structure) does not respect _pack_...
• #12347: DOC: typo in docstring numpy.random.beta, shape parameters must...
• #12349: TST, DOC: store circleci doc artifacts
• #12353: BUG: test, fix for threshold='nan'
• #12354: BUG: Fix segfault when an error occurs in np.fromfile
• #12355: BUG: fix a bug in npy_PyFile_Dup2 where it didn't return immediately...
• #12357: MAINT: Cleanup pavement file
• #12358: BUG: test, fix loading structured dtypes with padding
• #12362: MAINT: disable __array_function__ dispatch unless environment...
• #12363: MAINT: update gfortran RPATH for AIX/Windows non-support.
• #12364: NEP: clarify the purpose of "types" in __array_function__.
• #12366: MAINT: Refactor sorting header file
• #12372: BUG: random: Fix handling of a=0 for numpy.random.weibull.
• #12373: MAINT: Improve error message for legal but unsupported PEP3118...
• #12376: BUG: do not override exception on import failure
• #12377: NEP: move nep 15 from accepted to final
• #12378: TST: Update complex long double precision tests.
• #12380: BUG: Fix for #10533 np.dtype(ctype) does not respect endianness
• #12381: BUG: graceful DataSource __del__ when __init__ fails
• #12382: ENH: set correct __module__ for objects in numpy's public API
• #12388: ENH: allow arrays for start and stop in {lin,log,geom}space
• #12390: DEV: remove shim added in 1.4
• #12391: DEP: raise on a call to deprecated numpy.lib.function_base.unique
• #12392: DOC: Add release notes for ctypes improvements
• #12398: BUG: fix possible overlap issues with avx enabled
• #12399: DOC: Fix typo in polyint. Fixes #12386.
• #12405: ENH: Add support for np.dtype(ctypes.Union)
• #12407: BUG: Fall back to 'ascii' locale in build (if needed)
• #12408: BUG: multifield-view of MaskedArray gets bad fill_value
• #12409: MAINT: correct the dtype.descr docstring
• #12413: BUG: Do not double-quote arguments to the command line
• #12414: MAINT: Update cversion hash.
• #12417: BUG: Fix regression on np.dtype(ctypes.c_void_p)
• #12419: Fix PyArray_FillFunc function definitions
• #12420: gfortran needs -lpthread & -maix64(64 build) in AIX
• #12422: MNT: Reword error message about loading pickled data.
• #12424: BUG: Fix inconsistent cache keying in ndpointer
• #12429: MAINT: Update mailmap for 1.16.0 release.
• #12431: BUG/ENH: Fix use of ndpointer in return values
• #12437: MAINT: refactor datetime.c_metadata creation
• #12439: BUG: test, fix NPY_VISIBILITY_HIDDEN on gcc, which becomes NPY_NO_EXPORT
• #12440: BUG: don't override original errors when casting inside np.dot()...
• #12443: MAINT Use set litterals
• #12445: MAINT: Use list and dict comprehension when possible
• #12446: MAINT: Fixups to new functions in np.lib.recfunctions
• #12447: ENH: add back the multifield copy->view change
• #12448: MAINT: Review F401,F841,F842 flake8 errors (unused variables...
• #12455: TST: use condition directive for Azure 2.7 check
• #12458: MAINT, DOC: fix Azure README badge
• #12464: BUG: IndexError for empty list on structured MaskedArray.
• #12466: TST: use openblas for Windows CI
• #12470: MAINT: remove wrapper functions from numpy.core.multiarray
• #12471: ENH: override support for np.linspace and friends
• #12474: TST: enable dispatcher test coverage
• #12477: DOC: fix example for __call__. See #12451
• #12486: DOC: Update copyright year in the license
• #12488: ENH: implement matmul on NDArrayOperatorsMixin
• #12493: BUG: fix records.fromfile fails to read data >4 GB
• #12494: BUG: test, fix matmul, dot for vector array with stride[i]=0
• #12498: TST: sync Azure Win openblas
• #12501: MAINT: removed word/typo from comment in site.cfg.example
• #12556: BUG: only override vector size for avx code for 1.16
• #12562: DOC, MAINT: Make PYVER = 3 in doc/Makefile.
• #12563: DOC: more doc updates for structured arrays
• #12564: BUG: fix an unsafe PyTuple_GET_ITEM call
• #12565: Fix lgtm.com C/C++ build
• #12567: BUG: reorder operations for VS2015
• #12568: BUG: fix improper use of C-API
• #12569: BUG: Make new-lines in compiler error messages print to the console
• #12570: MAINT: don't check alignment size=0 arrays (RELAXED_STRIDES)
• #12573: BUG: fix refcount issue caused by #12524
• #12580: BUG: fix segfault in ctypeslib with obj being collected
• #12581: TST: activate shippable maintenance branches
• #12582: BUG: fix f2py pep338 execution method
• #12587: BUG: Make arr.ctypes.data hold a reference to the underlying...
• #12588: BUG: check for errors after PyArray_DESCR_REPLACE
• #12590: DOC, MAINT: Prepare for 1.16.0rc1 release.
• #12603: DOC: Fix markup in 1.16.0 release notes.
• #12621: BUG: longdouble with elsize 12 is never uint alignable.
• #12622: BUG: Add missing free in ufunc dealloc
• #12623: MAINT: add test for 12-byte alignment
• #12655: BUG: fix uint alignment asserts in lowlevel loops
• #12656: BENCH: don't fail at import time with old Numpy
• #12657: DOC: update 2018 -> 2019
• #12705: ENH: Better links in documentation
• #12706: MAINT: Further fixups to uint alignment checks
• #12707: BUG: Add 'sparc' to platforms implementing 16 byte reals.
• #12708: TST: Fix endianness in unstuctured_to_structured test
• #12710: TST: pin Azure brew version for stability.