1.14.0-changelog.rst

Contributors

A total of 100 people contributed to this release. People with a "+" by their names contributed a patch for the first time.

• Alexey Brodkin +
• Allan Haldane
• Andras Deak +
• Andrew Lawson +
• Anna Chiara +
• Antoine Pitrou
• Bernhard M. Wiedemann +
• Bob Eldering +
• Brandon Carter
• CJ Carey
• Charles Harris
• Chris Lamb
• Christoph Boeddeker +
• Christoph Gohlke
• Daniel Hrisca +
• Daniel Smith
• Danny Hermes
• David Freese
• David Hagen
• David Linke +
• David Schaefer +
• Dillon Niederhut +
• Egor Panfilov +
• Emilien Kofman
• Eric Wieser
• Erik Bray +
• Erik Quaeghebeur +
• Garry Polley +
• Gunjan +
• Han Shen +
• Henke Adolfsson +
• Hidehiro NAGAOKA +
• Hemil Desai +
• Hong Xu +
• Iryna Shcherbina +
• Jaime Fernandez
• James Bourbeau +
• Jamie Townsend +
• Jarrod Millman
• Jean Helie +
• Jeroen Demeyer +
• John Goetz +
• John Kirkham
• John Zwinck
• Jonathan Helmus
• Joseph Fox-Rabinovitz
• Joseph Paul Cohen +
• Joshua Leahy +
• Julian Taylor
• Jörg Döpfert +
• Keno Goertz +
• Kevin Sheppard +
• Kexuan Sun +
• Konrad Kapp +
• Kristofor Maynard +
• Licht Takeuchi +
• Loïc Estève
• Lukas Mericle +
• Marten van Kerkwijk
• Matheus Portela +
• Matthew Brett
• Matti Picus
• Michael Lamparski +
• Michael Odintsov +
• Michael Schnaitter +
• Michael Seifert
• Mike Nolta
• Nathaniel J. Smith
• Nelle Varoquaux +
• Nicholas Del Grosso +
• Nico Schlömer +
• Oleg Zabluda +
• Oleksandr Pavlyk
• Pauli Virtanen
• Pim de Haan +
• Ralf Gommers
• Robert T. McGibbon +
• Roland Kaufmann
• Sebastian Berg
• Serhiy Storchaka +
• Shitian Ni +
• Spencer Hill +
• Srinivas Reddy Thatiparthy +
• Stefan Winkler +
• Stephan Hoyer
• Steven Maude +
• SuperBo +
• Thomas Köppe +
• Toon Verstraelen
• Vedant Misra +
• Warren Weckesser
• Wirawan Purwanto +
• Yang Li +
• Ziyan Zhou +
• chaoyu3 +
• orbit-stabilizer +
• solarjoe
• wufangjie +
• xoviat +
• Élie Gouzien +

Pull requests merged

A total of 381 pull requests were merged for this release.

• #5580: BUG, DEP: Fix masked arrays to properly edit views. ( #5558 )
• #6053: MAINT: struct assignment "by field position", multi-field indices...
• #7994: BUG: Allow 'shape': () in __array_interface__ regardless of the...
• #8187: MAINT: Remove the unused keepdim argument from np.ufunc.accumulate
• #8278: MAINT: Make the refactor suggested in prepare_index
• #8557: ENH: add hermitian=False kwarg to np.linalg.matrix_rank
• #8722: DOC: Clarifying the meaning of small values for suppress print...
• #8827: BUG: Fix pinv for stacked matrices
• #8920: ENH: use caching memory allocator in more places
• #8934: MAINT: Use np.concatenate instead of np.vstack
• #8977: BUG: Fix all kinds of problems when itemsize == 0
• #8981: ENH: implement voidtype_repr and voidtype_str
• #8983: ENH: fix str/repr for 0d-arrays and int* scalars
• #9020: BUG: don't silence warnings in ufunc.reduce
• #9025: ENH: np.save() to align data at 64 bytes
• #9056: DOC: update structured array docs to reflect #6053
• #9065: DEP: 0 should be passed to bincount, not None
• #9083: MAINT: Improve error message from sorting with duplicate key
• #9089: MAINT: refine error message for __array_ufunc__ not implemented
• #9090: MAINT: Update master branch for 1.14.0 development.
• #9092: BUG remove memory leak in array ufunc override.
• #9096: ENH: Allow inplace also as keyword parameter for ndarray.byteswap
• #9099: TST: fix test_basic failure on Windows
• #9106: BUG: Array ufunc reduce out tuple
• #9110: BUG: Do not elide complex abs()
• #9112: BUG: ndarray.__pow__ does not check result of fast_scalar_power
• #9113: BUG: delay calls of array repr in getlimits
• #9115: BUG: Compilation crashes in MSVC when LIB or INCLUDE is not set
• #9116: DOC: link to stack from column_stack
• #9118: BUG: Fix reference count error of types when init multiarraymodule
• #9119: BUG: Fix error handling on PyCapsule when initializing multiarraymodule
• #9122: DOC: update 1.13 release note for MaskedArray, masked constants...
• #9132: DEP: Deprecate incorrect behavior of expand_dims.
• #9138: MAINT: Update .mailmap
• #9139: ENH: remove unneeded spaces in float/bool reprs, fixes 0d str
• #9141: DOC: Update ufunc documentation
• #9142: BUG: set default type for empty index array to numpy.intp
• #9149: DOC: Fix incorrect function signature in UFunc documentation.
• #9151: DOC: better link display text for Developer Zone.
• #9152: DOC: Fix some very minor spelling/grammar mistakes in docs
• #9155: MAINT: Take out code that will never be executed
• #9157: DOC: Fixed broken link to scipy developer zone
• #9164: BUG: have as_strided() keep custom dtypes
• #9167: BUG: ensure structured ndarray.__eq__,__ne__ defer when appropriate.
• #9168: MAINT: Simplify if statement
• #9174: BUG: allow pickling generic datetime
• #9176: DOC: Update protocols in git development document.
• #9181: COMPAT: PyPy calls clongdouble_int which raises a warning
• #9195: BUG: pull request 9087 modifies a tuple after use
• #9200: DOC: Update bincount docs to reflect gh-8348
• #9201: BUG: Fix unicode(unicode_array_0d) on python 2.7
• #9202: MAINT: Move ndarray.__str__ and ndarray.__repr__ to their own...
• #9205: DOC: Remove all references to bigndarray in documentation.
• #9209: ENH: Add an out argument to concatenate
• #9212: MAINT: Combine similar branches
• #9214: MAINT: Don't internally use the one-argument where
• #9215: BUG: Avoid bare except clauses
• #9217: BUG: handle resize of 0d array
• #9218: BUG: Only propagate TypeError from where we throw it
• #9219: DOC: Link to ufunc.outer from np.outer
• #9220: MAINT: Factor out code duplicated by nanmedian and nanpercentile
• #9226: DOC, ENH: Add 1.13.0-changelog.rst
• #9238: DOC: BLD: fix lots of Sphinx warnings/errors.
• #9241: MAINT: Fixup release notes, changelogs after #9238 merge.
• #9242: BUG: Make 0-length dim handling of tensordot consistent with...
• #9246: ENH: Release the GIL in einsum() special-cased loops
• #9247: BUG: fix missing keyword rename for common block in numpy.f2py
• #9253: DOC: Add isnat/positive ufunc to documentation.
• #9259: MAINT: Use XOR for bool arrays in np.diff
• #9260: BUG: don't elide into readonly and updateifcopy temporaries
• #9264: DOC: some doc build maintenance and f2py doc updates
• #9266: BUG: Fix unused variable in ufunc_object.c,
• #9268: ENH: testing: load available nose plugins that are external to...
• #9271: BUG: fix issue when using python setup.py somecommand --force.
• #9280: BUG: Make extensions compilable with MinGW on Py2.7
• #9281: DOC: add @ operator in array vs. matrix comparison doc
• #9285: BUG: Fix Intel compilation on Unix.
• #9292: MAINT: Fix lgtm alerts
• #9294: BUG: Fixes histogram monotonicity check for unsigned bin values
• #9300: BUG: PyArray_CountNonzero does not check for exceptions
• #9302: BUG: Fix fillvalue
• #9306: BUG: f2py: Convert some error messages printed to stderr to exceptions.
• #9310: BUG: fix wrong ndim used in empty where check
• #9316: BUG: runtest -t should recognize development mode
• #9320: DOC: Use x1 and x2 in the heaviside docstring.
• #9322: BUG: np.ma.astype fails on structured types
• #9323: DOC: Add $PARAMS to the isnat docstring
• #9324: DOC: Fix missing asterisks in git development_setup doc page
• #9325: DOC: add a NumFOCUS badge to README.md
• #9332: ENH: fix 0d array printing using str or formatter.
• #9335: BUG: umath: un-break ufunc where= when no out= is given
• #9336: BUG: Fix various problems with the np.ma.masked constant
• #9337: BUG: Prevent crash if ufunc doc string is null
• #9341: BUG: np.resize discards empty shapes
• #9343: BUG: recfunctions fail in a bunch of ways due to using .descr
• #9344: DOC: fixes issue #9326, by removing the statement.
• #9346: BUG: void masked fillvalue cannot be cast to void in python 3
• #9354: BUG: Prevent hang traversing ufunc userloop linked list
• #9357: DOC: Add examples for complex dtypes
• #9361: DOC: isscalar add example for str
• #9362: ENH: Rearrange testing module to isolate nose dependency.
• #9364: BUG: ')' is printed at the end pointer of the buffer in numpy.f2py.
• #9369: BUG: fix error in fromstring function from numpy.core.records
• #9375: DOC: Document the internal workings of PY_ARRAY_UNIQUE_SYMBOL
• #9380: DOC: Forward port 1.13.1 notes and changelog.
• #9381: TST: test doc string of COMMON block arrays for numpy.f2py.
• #9387: MAINT: Simplify code using PyArray_ISBYTESWAPPED macro.
• #9388: MAINT: Use PyArray_ISBYTESWAPPED instead of !PyArray_ISNOTSWAPPED.
• #9389: DOC: Fix reference, PyArray_DescrNew -> PyArray_NewFromDescr
• #9392: DOC: UPDATEIFCOPY raises an error if not an array.
• #9399: DOC: document how to free memory from PyArray_IntpConverter.
• #9400: MAINT: Further unify handling of unnamed ufuncs
• #9403: MAINT: Replace tab escapes with four spaces
• #9407: DOC: add suppress_warnings to the testing routine listing.
• #9408: BUG: various fixes to np.gradient
• #9411: MAINT/BUG: improve gradient dtype handling
• #9412: BUG: Check for exception in sort functions
• #9422: DOC: correct formatting of basic.types.html
• #9423: MAINT: change http to https for numfocus.org link in README
• #9425: ENH: Einsum calls BLAS if it advantageous to do so
• #9426: DOC: Add a link to einsum_path
• #9431: ENH: distutils: make msvc + mingw-gfortran work
• #9432: BUG: Fix loss of masks in masked 0d methods
• #9433: BUG: make np.transpose return a view of the mask
• #9434: MAINT: Remove unittest dependencies
• #9437: DOC: Update 1.14.0 release notes.
• #9446: BUG: Inlined functions must be defined somewhere.
• #9447: API: Make a.flat.__array__ return a copy when a non-contiguous.
• #9452: MAINT: Use new-style classes on 2.7
• #9454: MAINT: Remove branch in __array__ where if and else were the...
• #9457: MAINT: Add a common subclass to all the masked ufunc wrappers
• #9458: MAINT: Improve performance of np.copyto(where=scalar)
• #9469: BUG: Fix true_divide when dtype=np.float64 specified.
• #9470: MAINT: Make setxor1d a bit clearer and speed it up
• #9471: BLD: remove -xhost flag from IntelFCompiler.
• #9475: DEP: deprecate rollaxis
• #9482: MAINT: Make diff iterative instead of recursive
• #9487: DEP: Letting fromstring pretend to be frombuffer is a bad idea
• #9490: DOC: Replace xrange by range in quickstart docs
• #9491: TST: Add filter for new Py3K warning in python 2
• #9492: ENH: Add np.polynomial.chebyshev.chebinterpolate function.
• #9498: DOC: fix versionadded in docstring for moveaxis
• #9499: MAINT/BUG: Improve error messages for dtype reassigment, fix...
• #9503: MAINT: Move variables into deepest relevant scope, for clarity
• #9505: BUG: issubdtype is inconsistent on types and dtypes
• #9517: MAINT/DOC: Use builtin when np.{x} is builtins.{x}.
• #9519: MAINT: Remove level= keyword from test arguments.
• #9520: MAINT: types.TypeType does not ever need to be used
• #9521: BUG: Make issubclass(np.number, numbers.Number) return true
• #9522: BUG: Fix problems with obj2sctype
• #9524: TST, MAINT: Add __init__.py files to tests directories.
• #9527: BUG: Fix scalar methods to receive keyword arguments
• #9529: BUG: The NAT deprecation warning should not be given for every...
• #9536: ENH: Show domain and window as kwargs in repr
• #9540: BUG: MaskedArray _optinfo dictionary is not updated when calling...
• #9543: DOC: Adding backslash between double-backtick and s.
• #9544: MAINT: Use the error_converting macro where possible
• #9545: DEP: Deprecate the event argument to datetime types, which is...
• #9550: DOC: removes broken docstring example (source code, png, pdf)...
• #9552: DOC, BUG: Fix Python 3.6 invalid escape sequence.
• #9554: BUG: fix regression in 1.13.x in distutils.mingw32ccompiler.
• #9564: BUG: fix distutils/cpuinfo.py:getoutput()
• #9574: BUG: deal with broken hypot() for MSVC on win32
• #9575: BUG: deal with broken cabs*() for MSVC on win32
• #9577: BUG: Missing dirichlet input validation
• #9581: DOC: Fix link in numpy.ndarray.copy method (missing backticks)
• #9582: ENH: Warn to change lstsq default for rcond
• #9586: DOC: update example in np.nonzero docstring
• #9588: MAINT: Remove direct access to flatiter attributes
• #9590: ENH: Remove unnecessary restriction in noncen-f
• #9591: MAINT: Remove unnecessary imports
• #9599: BUG: fix infinite loop when creating np.pad on an empty array
• #9601: DOC: rot90 wrongly positioned versionadded directive.
• #9604: MAINT: Refactor the code used to compute sha256, md5 hashes
• #9606: MAINT: Remove global statement in linalg.py
• #9609: BUG: Add __ne__ method to dummy_ctype class.
• #9610: BUG: core: fix wrong method flags for scalartypes.c.src:gentype_copy
• #9611: MAINT: remove try..except clause.
• #9613: DOC: Update release notes for noncentral_f changes.
• #9614: MAINT: Fix a comment regarding the formula for arange length
• #9618: DOC: Fix type definitions in mtrand
• #9619: ENH: Allow Fortran arrays of dimension 0
• #9624: BUG: memory leak in np.dot of size 0
• #9626: BUG: Fix broken runtests '-t' option.
• #9629: BUG: test, fix issue #9620 __radd__ in char scalars
• #9630: DOC: Updates order of parameters in save docstring
• #9636: MAINT: Fix compiler warnings and update travis jobs
• #9638: BUG: ensure consistent result dtype of count_nonzero
• #9639: MAINT: Refactor updateifcopy
• #9640: BUG: fix padding an empty array in reflect mode.
• #9643: DOC: add new steering council members.
• #9645: ENH: enable OpenBLAS on windows.
• #9648: DOC: Correct the signature in pad doc for callable mode.
• #9649: DOC: Fixed doc example of apply along axis with 3D return
• #9652: BUG: Make system_info output reproducible
• #9658: BUG: Fix usage of keyword "from" as argument name for "can_cast".
• #9667: MAINT: Simplify block implementation
• #9668: DOC: clarify wording in tutorial
• #9672: BUG: dot/matmul 'out' arg should accept any ndarray subclass
• #9681: MAINT: Add block benchmarks
• #9682: DOC: Add whitespace after "versionadded::" directive so it actually...
• #9683: DOC: Add polyutils subpackage to reference documentation
• #9685: BUG: Fixes #7395, operator.index now fails on numpy.bool_
• #9688: MAINT: rework recursive guard to keep array2string signature
• #9691: PEP 3141 numbers should be considered scalars
• #9692: ENH: Add support of ARC architecture
• #9695: DOC: start is not needed even when step is given.
• #9700: DOC: Add mandatory memo argument to __deepcopy__ method documentation
• #9701: DOC: Add keepdims argument for ndarray.max documentation
• #9702: DOC: Warn about the difference between np.remainder and math.remainder
• #9703: DOC: Fix mistaken word in nanprod docstring
• #9707: MAINT: When linspace's step is a NumPy scalar, do multiplication in-place
• #9709: DOC: allclose doesn't require matching shapes
• #9711: BUG: Make scalar function elision check if writeable.
• #9715: MAINT: Fix typo "Porland" -> "Portland" in building doc.
• #9718: DEP: Deprecate truth testing on empty arrays
• #9720: MAINT: Remove unnecessary special-casing of scalars in isclose
• #9724: BUG: adjust gfortran version search regex
• #9725: MAINT: cleanup circular import b/w arrayprint.py,numeric.py
• #9726: ENH: Better error message for savetxt when X.ndim > 2 or X.ndim...
• #9737: MAINT: Use zip, not enumerate
• #9740: BUG: Ensure _npy_scaled_cexp{,f,l} is defined when needed.
• #9741: BUG: core: use npy_cabs for abs() for np.complex* scalar types
• #9743: MAINT: Use PyArray_CHKFLAGS in more places.
• #9749: BUG: Fix loss of precision for large values in long double divmod
• #9752: BUG: Errors thrown by 0d arrays in setitem are silenced and replaced
• #9753: DOC: Fix ndarray.__setstate__ documentation, it only takes one...
• #9755: BUG: Cython 0.27 breaks NumPy on Python 3.
• #9756: BUG/TST: Check if precision is lost in longcomplex
• #9762: MAINT: Use the PyArray_(GET|SET)_ITEM functions where possible
• #9768: MAINT: Cleanup ma.array.__str__
• #9770: MAINT,BUG: Fix mtrand for Cython 0.27.
• #9773: BUG: Fixes optimal einsum path for multi-term intermediates
• #9778: BUG: can_cast(127, np.int8) is False
• #9779: BUG: np.ma.trace gives the wrong result on ND arrays
• #9780: MAINT: Make f2py generated file not contain the (local) date.
• #9782: DOC: Update after NumPy 1.13.2 release.
• #9784: BUG: remove voidtype-repr recursion in scalartypes.c/arrayprint.py
• #9785: BUG: Fix size-checking in masked_where, and structured shrink_mask
• #9792: ENH: Various improvements to Maskedarray repr
• #9796: TST: linalg: add basic smoketest for cholesky
• #9800: DOC: Clean up README
• #9803: DOC: add missing underscore in set_printoptions
• #9805: CI: set correct test mode for appveyor
• #9806: MAINT: Add appveyor badge to README
• #9807: MAINT: Make appveyor config a dot-file
• #9810: DOC: Improve ndarray.shape documentation.
• #9812: DOC: update scipy.integrate recommendation
• #9814: BUG: Fix datetime->string conversion
• #9815: BUG: fix stray comma in _array2string
• #9817: BUG: Added exception for casting numpy.ma.masked to long
• #9822: BUG: Allow subclasses of MaskedConstant to behave as unique singletons
• #9824: BUG: Fixes for np.random.zipf
• #9826: DOC: Add unravel_index examples to np.arg(min|max|sort)
• #9828: DOC: Improve documentation of axis parameter in numpy.unpackbits()
• #9835: BENCH: Added missing ufunc benchmarks
• #9840: DOC: ndarray.__copy__ takes no arguments
• #9842: BUG: Prevent invalid array shapes in seed
• #9845: DOC: Refine SVD documentation
• #9849: MAINT: Fix all special-casing of dtypes in count_nonzero
• #9854: BLD: distutils: auto-find vcpkg include and library directories
• #9856: BUG: Make bool(void_scalar) and void_scalar.astype(bool) consistent
• #9858: DOC: Some minor fixes regarding import_array
• #9862: BUG: Restore the environment variables when import multiarray...
• #9863: ENH: Save to ZIP files without using temporary files.
• #9865: DOC: Replace PyFITS reference with Astropy and PyTables with...
• #9866: BUG: Fix runtests --benchmark-compare in python 3
• #9868: DOC: Update arraypad to use np.pad in examples
• #9869: DOC: Make qr options render correctly as list.
• #9881: BUG: count_nonzero treats empty axis tuples strangely
• #9883: ENH: Implement ndarray.__format__ for 0d arrays
• #9884: BUG: Allow unravel_index(0, ()) to return ()
• #9887: BUG: add.reduce gives wrong results for arrays with funny strides
• #9888: MAINT: Remove workarounds for gh-9527
• #9889: MAINT: Tidy np.histogram, and improve error messages
• #9893: ENH: Added compatibility for the NAG Fortran compiler, nagfor
• #9896: DOC: Unindent enumeration in savetxt docstring
• #9899: Remove unused isscalar imports, and incorrect documentation using...
• #9900: MAINT/BUG: Remove special-casing for 0d arrays, now that indexing...
• #9904: MAINT: Make warnings for nanmin and nanmax consistent
• #9911: CI: travis: switch to container
• #9912: BENCH: histogramming benchmarks
• #9913: MAINT: Tidy up Maskedarray repr
• #9916: DOC: Clarify behavior of genfromtxt names field
• #9920: DOC: dot: Add explanation in case b has only 1 dimension.
• #9925: DOC: ndarray.reshape allows shape as int arguments or tuple
• #9930: MAINT: Add parameter checks to polynomial integration functions.
• #9936: DOC: Clarify docstring for numpy.array_split
• #9941: ENH: Use Dragon4 algorithm to print floating values
• #9942: ENH: Add PGI flang compiler support for Windows
• #9944: MAINT/BUG: Don't squash useful error messages in favor of generic...
• #9945: DOC: fix operation plural in along axis glossary
• #9946: DOC: describe the expansion of take and apply_along_axis in detail
• #9947: MAINT/TST: Tidy dtype indexing
• #9950: BUG: Passing an incorrect type to dtype.__getitem__ should raise...
• #9952: ENH: add Decimal support to numpy.lib.financial
• #9953: MAINT: Add a PyDataType_ISUNSIZED macro
• #9957: DOC: update asv url
• #9961: BUG: Allow float64('1e10000') to overflow
• #9962: MAINT: Rename formatters to match scalar type names
• #9965: BLD: Disable npymath whole program opt (LTCG) on win32
• #9966: BUG: str(np.float) should print with the same number of digits...
• #9967: MAINT: Separate correct longdouble.__float__ from incorrect...
• #9971: BUG: Fix casting from longdouble to long
• #9973: TST: Fix error in test on PyPy, add comment explaining known...
• #9976: BUG: Ensure lstsq can handle RHS with all sizes.
• #9977: MAINT: distutils: trivial cleanups
• #9978: BUG: cast to str_ should not convert to pure-python intermediate
• #9983: ENH: let f2py discover location of libgfortran
• #9985: ENH: skip NPY_ALLOW_C_API for UFUNC_ERR_IGNORE
• #9986: MAINT: Remove similar branches from linalg.lstsq
• #9991: MAINT: small robustness change for mingw support on Windows.
• #9994: BUG: test was not using 'mode'
• #9996: ENH: Adding order= keyword to np.eye().
• #9997: BUG: prototypes for [cz]dot[uc] are incorrect
• #9999: ENH: Make np.in1d() work for unorderable object arrays
• #10000: MAINT: Fix test_int_from_huge_longdouble on Darwin.
• #10005: DOC: reword PyArray_DiscardWritebackIfCopy description
• #10006: NEP: Drop Python2 support.
• #10007: MAINT: simplify logic from #9983
• #10008: MAINT: Backcompat fixes for dragon4 changes
• #10011: TST: Group together all the nested_iter tests
• #10017: REV: Undo bad rebase in 7fdfdd6a52fc0761c0d45931247c5ed2480224eb...
• #10021: ENH: Don't show the boolean dtype in array_repr
• #10022: MAINT: Update c-api version and hash for NumPy 1.14.
• #10030: MAINT: Legacy mode specified as string, fix all-zeros legacy...
• #10031: BUG: Fix f2py string variables in callbacks.
• #10032: MAINT: Remove newline before dtype in repr of arrays
• #10034: MAINT: legacy-printing-mode preserves 1.13 float & complex str
• #10042: BUG: Allow int to be called on nested object arrays, fix np.str_.__int__
• #10044: DEP: FutureWarning for void.item(): Will return bytes
• #10049: DOC: Add copy of deprecated defindex.html template.
• #10052: BUG: Fix legacy printing mode check.
• #10053: STY: C style whitespace fixups
• #10054: ENH: Add encoding option to numpy text IO.
• #10055: BUG: Changed dump(a, F) so it would close file
• #10057: DOC: v/h/dstack docstr shouldn't imply deprecation
• #10065: DOC, BLD: Update site.cfg.example on the MKL part.
• #10067: MAINT: Replace sphinx extension sphinx.ext.pngmath by sphinx.ext.imgmath.
• #10068: BUG: Fix memory leak for subclass slicing
• #10072: MAINT: Fix minor typos in numpy/core/fromnumeric.py
• #10079: DOC: mention generalized ufuncs, document signature attribute
• #10096: BUG: Fix assert_equal on time-like objects
• #10097: BUG: Fix crash for 0d timedelta repr
• #10101: BUG: Fix out-of-bounds access when handling rank-zero ndarrays.
• #10105: DOC: Update license documentation.
• #10108: DOC: Add documentation for datetime_data
• #10109: DOC: fix the lack of np.
• #10111: ENH: Improve alignment of datetime64 arrays containing NaT
• #10112: MAINT: Simplify IntegerFormatter
• #10113: BUG: Fix further out-of-bounds accesses when handling 0d ndarrays
• #10114: MAINT: Remove duplicate cond check from assert_array_compare
• #10116: BLD: [ipo] compilation error with intel compiler
• #10120: BUG: stray comma should be preserved for legacy printing
• #10121: DOC: Summarize printing changes in release notes
• #10125: BLD: Add license file to NumPy wheels.
• #10129: ENH: Strip trailing spaces from continuation in multiline arrayprint
• #10130: MAINT: Simplify _leading_trailing
• #10131: BUG: Fix downcasting in _array2string
• #10136: BUG: edgeitems kwarg is ignored
• #10143: MAINT: Combine legacy sections of _formatArray
• #10159: DOC: Update 1.14 notes
• #10160: BUG: test, fix problems from PR #9639
• #10164: MAINT/BUG: Simplify _formatArray, fixing array_repr(matrix) in...
• #10166: DOC: document PyArray_ResolveWritebackIfCopy
• #10168: DOC: continuation of PyArray_ResolveIfCopy fixes
• #10172: BUG: The last line of formatArray is not always wrapped correctly
• #10175: BUG: linewidth was not respected for arrays other than 1d
• #10176: ENH: add suffix option to array2str, wraps properly
• #10177: MAINT, BUG: Final 1.14 formatting fixes
• #10182: BUG: Extra space is inserted on first line for long elements
• #10190: BUG: Fix regression in np.ma.load in gh-10055
• #10200: BUG: Ufunc reduce reference leak (backport)
• #10202: BUG: Fix bugs found by testing in release mode.
• #10272: BUG: Align extra-dll folder name with auditwheel
• #10275: BUG: fix duplicate message print
• #10276: MAINT: Workaround for new travis sdist failures.
• #10311: BUG: Make sure einsum default value of optimize is True.
• #10312: BUG: Handle NaNs correctly in arange
• #10313: BUG: Don't reimplement isclose in np.ma
• #10315: DOC: NumPy 1.14.0 release prep.