1.13.0-changelog.rst

Changelog

Contributors

A total of 102 people contributed to this release. People with a "+" by their names contributed a patch for the first time.

• 1. Jesse Jiryu Davis +
• Alessandro Pietro Bardelli +
• Alex Rothberg +
• Alexander Shadchin
• Allan Haldane
• Andres Guzman-Ballen +
• Antoine Pitrou
• Antony Lee
• B R S Recht +
• Baurzhan Muftakhidinov +
• Ben Rowland
• Benda Xu +
• Blake Griffith
• Bradley Wogsland +
• Brandon Carter +
• CJ Carey
• Charles Harris
• Christoph Gohlke
• Danny Hermes +
• David Hagen +
• Duke Vijitbenjaronk +
• Egor Klenin +
• Elliott Forney +
• Elliott M Forney +
• Endolith
• Eric Wieser
• Erik M. Bray
• Eugene +
• Evan Limanto +
• Felix Berkenkamp +
• François Bissey +
• Frederic Bastien
• Greg Young
• Gregory R. Lee
• Importance of Being Ernest +
• Jaime Fernandez
• Jakub Wilk +
• James Cowgill +
• James Sanders
• Jean Utke +
• Jesse Thoren +
• Jim Crist +
• Joerg Behrmann +
• John Kirkham
• Jonathan Helmus
• Jonathan L Long
• Jonathan Tammo Siebert +
• Joseph Fox-Rabinovitz
• Joshua Loyal +
• Juan Nunez-Iglesias +
• Julian Taylor
• Kirill Balunov +
• Likhith Chitneni +
• Loïc Estève
• Mads Ohm Larsen
• Marein Könings +
• Marten van Kerkwijk
• Martin Thoma
• Martino Sorbaro +
• Marvin Schmidt +
• Matthew Brett
• Matthias Bussonnier +
• Matthias C. M. Troffaes +
• Matti Picus
• Michael Seifert
• Mikhail Pak +
• Mortada Mehyar
• Nathaniel J. Smith
• Nick Papior
• Oscar Villellas +
• Pauli Virtanen
• Pavel Potocek
• Pete Peeradej Tanruangporn +
• Philipp A +
• Ralf Gommers
• Robert Kern
• Roland Kaufmann +
• Ronan Lamy
• Sami Salonen +
• Sanchez Gonzalez Alvaro
• Sebastian Berg
• Shota Kawabuchi
• Simon Gibbons
• Stefan Otte
• Stefan Peterson +
• Stephan Hoyer
• Søren Fuglede Jørgensen +
• Takuya Akiba
• Tom Boyd +
• Ville Skyttä +
• Warren Weckesser
• Wendell Smith
• Yu Feng
• Zixu Zhao +
• Zè Vinícius +
• aha66 +
• davidjn +
• drabach +
• drlvk +
• jsh9 +
• solarjoe +
• zengi +

Pull requests merged

A total of 309 pull requests were merged for this release.

• #3861: ENH: Make it possible to NpyIter_RemoveAxis an empty dimension
• #5302: Fixed meshgrid to return arrays with same dtype as arguments.
• #5726: BUG, API: np.random.multivariate_normal behavior with bad covariance...
• #6632: TST/BUG: fromfile - fix test and expose bug with io class argument
• #6659: BUG: Let linspace accept input that has an array_interface.
• #7742: Add axis argument to numpy.unique
• #7862: BLD: rewrite np.distutils.exec_command.exec_command()
• #7997: ENH: avoid temporary arrays in expressions (again)
• #8043: ENH: umath: ensure ufuncs are well-defined with memory overlapping...
• #8106: DOC: Document release procedure with a walkthrough.
• #8194: BUG: np.piecewise not working for scalars
• #8235: BUG: add checks for some invalid structured dtypes. Fixes #2865.
• #8241: MAINT: Prepare for 1.13.0 after 1.12.x branch
• #8242: BUG: Update operator ** tests for new behavior.
• #8244: DOC: fix typos in arrayprint docstrings.
• #8247: ENH: Add __array_ufunc__
• #8251: MAINT: Cleaned up mailmap
• #8267: DOC: Changed shape assignment example to reshape. Elaborated...
• #8271: TST: Path test should resolve symlinks when comparing
• #8277: DOC: improve comment in prepare_index
• #8279: BUG: bool(dtype) is True
• #8281: DOC: Update 1.12.0 release notes.
• #8284: BUG: Fix iteration over reversed subspaces in mapiter@name@
• #8285: BUG: Fix pavement.py write_release_task.
• #8287: DOC: Update 1.13.0 release notes.
• #8290: MAINT: let average preserve subclass information.
• #8297: DEP: Handle expired deprecations.
• #8299: BUG: Make f2py respect kind specifications for real parameters
• #8302: BUG: Fix PyPy crash in PyUFunc_GenericReduction.
• #8308: BUG: do not link to Accelerate if OpenBLAS, MKL or BLIS are found.
• #8312: DEP: Drop deprecated boolean indexing behavior and update to...
• #8318: BLD: blacklist powl (longdouble power function) on OS X.
• #8326: ENH: Vectorize packbits with SSE2
• #8327: BUG: Fix packbits to correctly handle empty arrays
• #8335: BUG: Fix ndarray.tofile large file corruption in append mode
• #8337: BUG: fix test_api test that fails intermittently in python 3
• #8343: TST: Ellipsis indexing creates a view
• #8348: ENH: Allow bincount(..., minlength=0).
• #8349: BUG: Apply more robust string converts in loadtxt
• #8351: BUG: correct letter case
• #8354: BUG: Fix suppress_warnings (again) for Python 3.6.
• #8355: Fix building extensions with MinGW for Python 3.5
• #8356: Allow extensions to be built with MinGW in a virtualenv
• #8360: MAINT: Drop special case code for python2 < 2.7 and python3 <...
• #8364: BUG: handle unmasked NaN in ma.median like normal median
• #8366: BUG: fix nanpercentile not returning scalar with axis argument
• #8367: xlocale.h is not available in newlib / Cygwin
• #8368: ENH: Implement most linalg operations for 0x0 matrices
• #8369: TST: Fix various incorrect linalg tests
• #8374: DOC: Fixed minor typo in William Gosset's name
• #8377: Switch to the PyPI version of plex to generate lapack_lite
• #8380: DOC: Update 1.12.0-notes.rst.
• #8381: MAINT: Rebuild lapack lite
• #8382: DEP: Fix escaped string characters deprecated in Python 3.6.
• #8384: ENH: Add tool to check for deprecated escaped characters.
• #8388: API: Return scalars for scalar inputs to np.real/imag
• #8389: ENH: retune apply_along_axis nanmedian cutoff
• #8395: DOC: create 1.11.3 release notes.
• #8398: BUG: Fix author search in announce.py
• #8400: Fix corrcoef and cov rowvar param handling
• #8401: DOC, MAINT: Update 1.12.0 notes and mailmap.
• #8410: BUG: Fixed behavior of assert_array_less for +/-inf
• #8414: BUG: fixed failure of np.ma.median for 1-D even arrays.
• #8416: BUG operations involving MaskedArray with output given do not...
• #8421: ENH: Add isnat function and make comparison tests NAT specific
• #8423: Adding isin function for multidimensional arrays
• #8426: BUG: Fix apply_along_axis() for when func1d() returns a non-ndarray
• #8434: TST: Update 3.6-dev tests to 3.6 after Python final release.
• #8441: BUG: Fix crash on 0d return value in apply_along_axis
• #8443: BUG: fix set memmap offset attribute correctly when offset is...
• #8445: BUG: correct norm='ortho' scaling for rfft when n != None
• #8446: ENH: gradient support for unevenly spaced data
• #8448: TST: remove a duplicate test. Closes gh-8447.
• #8452: BUG: assert_almost_equal fails on subclasses that cannot handle...
• #8454: MAINT: Fix building extensions with MinGW in WinPython 3.4
• #8464: [DOC]Small release doc fix
• #8468: BUG: Ensure inf/nan removal in assert_array_compare is matrix-safe.
• #8470: DOC: Add example to np.savez_compressed
• #8474: MAINT: use env in shebang instead of absolute path to python
• #8475: DOC: improve clip docstring
• #8478: MAINT: Forward port accumulated changes from the 1.12.0 release.
• #8482: TST: switch to ubuntu yakkety for i386 testing
• #8483: BUG: fix wrong future nat warning and equiv type logic error
• #8486: BUG: Prevent crash for length-0 input to fromrecords
• #8488: ENH: Improve the alignment of recarray.__repr__
• #8489: BUG: fix wrong masked median for some special cases
• #8490: DOC: Place np.average in inline code
• #8491: TST: work around isfinite inconsistency on i386
• #8494: BUG: guard against replacing constants without '_' spec
• #8496: Update LICENSE.txt to 2017
• #8497: BUG: Fix creating a np.matrix from string syntax involving booleans
• #8501: Changing spurious Legendre reference to Chebyshev in chebfit...
• #8504: ENH: hard-code finfo parameters for known types
• #8508: BUG: Fix loss of dimensionality of np.ma.masked in ufunc
• #8524: BUG: fix mean for float 16 non-array inputs
• #8527: DOC: fix return value for PyArray_Resize
• #8539: BUG: core: in dot(), make copies if out has memory overlap with...
• #8540: DOC: Update arrays.ndarray.rst
• #8541: DOC: Revert 8540 patch 1
• #8542: MAINT: typo in histogram docstring
• #8551: DOC: Missing backticks
• #8555: Fixing docstring error in polyvander2d
• #8558: DOC: Improve documentation of None as interval bounds in clip.
• #8567: TST: core: use aligned memory for dot() out= arrays
• #8568: TST: re-enable PPC longdouble spacing tests
• #8569: ENH: Add missing __tracebackhide__ to testing functions.
• #8570: BUG: fix issue #8250 when np.array gets called on an invalid...
• #8571: BUG: fix calling python api with error set and minor leaks
• #8572: MAINT: remove ma out= workaround
• #8575: DOC: fix several typos #8537.
• #8584: MAINT: Use the same exception for all bad axis requests
• #8586: MAINT: PyPy3 compatibility: sys.getsizeof()
• #8590: BUG MaskedArray __eq__ wrong for masked scalar, multi-d recarray
• #8591: BUG: make np.squeeze always return an array, never a scalar
• #8592: MAINT: Remove __setslice__ and __getslice__
• #8594: BUG: Fix MaskedArray.__setitem__
• #8596: BUG: match hard-coded finfo to calculated MachAr
• #8602: BUG: Make iscomplexobj compatible with custom dtypes again
• #8605: DOC: gradient uses 1st order central difference in the interior
• #8606: Revert "DOC: gradient uses 1st order central difference in the...
• #8610: Revert "BUG: make np.squeeze always return an array, never a...
• #8611: DOC: The axis argument of average can be a tuple of ints
• #8612: MAINT: Decrease merge conflicts in release notes
• #8614: BUG: Don't leak internal exceptions when given an empty array
• #8617: BUG: Copy meshgrid after broadcasting
• #8618: BUG: Fix undefined behaviour induced by bad __array_wrap__
• #8619: BUG: blas_info should record include_dirs
• #8625: DOC: Create 1.12.1 release notes.
• #8629: ENH: Improve the efficiency of indices
• #8631: Fix typo in fill_diagonal docstring.
• #8633: DOC: Mention boolean arrays in the ix documentation.
• #8636: MAINT: ensure benchmark suite is importable on old numpy versions
• #8638: BUG: fix wrong odd determination in packbits
• #8643: BUG: Fix double-wrapping of object scalars
• #8645: MAINT: Use getmask where possible
• #8646: ENH: Allow for an in-place nan_to_num conversion
• #8647: Fix various bugs in np.ma.where
• #8649: Upgrade to Lapack lite 3.2.2
• #8650: DOC: Fix obsolete data in readme
• #8651: MAINT: Split lapack_lite more logically across files
• #8652: TST: Improve testing of read-only mmaps
• #8655: MAINT: Squelch parenthesis warnings from GCC
• #8656: BUG: allow for precision > 17 in longdouble repr test
• #8658: BUG: fix denormal linspace test for longdouble
• #8659: BUG: PPC64el machines are POWER for Fortran
• #8663: ENH: Fix alignment of repr for array subclasses
• #8665: BUG: Look up methods on MaskedArray in _frommethod
• #8667: BUG: Preserve identity of dtypes in make_mask_descr
• #8668: DOC: Add more examples for np.c_
• #8669: MAINT: Warn users when calling np.ma.MaskedArray.partition function.
• #8672: BUG: Use int for axes, not intp
• #8674: BUG: Remove extra digit in binary_repr at limit
• #8675: BUG: Fix problems detecting runtime for MSYS2 compiler on Windows
• #8677: MAINT: We can now rely on itertools.izip_longest existing
• #8678: BUG: Fix argsort vs sort in Masked arrays
• #8680: DOC: Removed broken link
• #8682: ENH: allow argument to matrix_rank to be stacked
• #8685: ENH: add dtype.ndim
• #8688: DOC: Added note to np.diff
• #8692: MAINT: Fix deprecated escape sequences
• #8694: BUG: missing comma disabled some header checks
• #8695: MAINT: Remove numpy-macosx-installer and win32build directories.
• #8698: DOC: fix incorrect mask value when value was changed
• #8702: DOC: Fixed small mistakes in numpy.copy documentation.
• #8704: BUG: Fix deepcopy regression for empty arrays.
• #8705: BUG: fix ma.median for empty ndarrays
• #8709: DOC: Fixed minor typos in temp_elide.c
• #8713: BUG: Don't signal FP exceptions in np.absolute
• #8716: MAINT: Mark some tests with slow decorator
• #8718: BUG: Fix assert statements in random.choice tests
• #8729: DOC: Add float_power to routines.math documentation autosummary
• #8731: DOC: added linalg.multi_dot to doc
• #8737: DOC: Mention that expand_dims and squeeze are inverses
• #8744: MAINT: Remove files and constants that were only needed for Bento.
• #8745: TST: Remove unused env from tox
• #8746: DOC: Update 1.12.1 release notes.
• #8749: DOC: Add 1.12.1 release notes to documentation.
• #8750: BUG: Fix np.average for object arrays
• #8754: ENH: Allows building npy_math with static inlining
• #8756: BUG: Correct lapack ld* args
• #8759: BUG: Add HOME to the git environment.
• #8761: MAINT: better warning message when running build_src from sdist
• #8762: BUG: Prevent crash in poly1d.__eq__
• #8781: BUG: Revert gh-8570.
• #8788: BUG: Fix scipy incompatibility with cleanup to poly1d
• #8792: DOC: Fix typos
• #8793: DOC: fix minor docstring typos
• #8795: ENH: Add the 'heaviside' ufunc.
• #8796: BUG: fix regex of determineexprtype_re_3 in numpy/f2py/crackfortran.py
• #8799: DOC: Include np. prefix in meshgrid examples
• #8801: BUG: fix the error msg of empty hstack input
• #8806: BUG: Raise TypeError on ternary power
• #8807: TST: Prove that poly1d coeffs are immutable
• #8813: MAINT: tidy up some of npyio
• #8816: BUG: np.lib.index_tricks.r_ mutates its own state
• #8820: DOC: Add 'heaviside' to the ufunc documentation.
• #8822: DOC: Use gray and hsv colormaps in examples
• #8824: MAINT: a couple distutils cleanups
• #8825: STY: Fix bad style in umath_linalg
• #8828: DOC: Add missing release note for #8584
• #8830: DOC: added a whitespace so that sphinx directive displays correctly
• #8832: MAINT: Remove python <2.7,<3.3 string/unicode workarounds
• #8834: BENCH: use initialized memory for count_nonzero benchmark
• #8835: DOC: Include nextafter and spacing function in documentation.
• #8836: DOC: Several documentation fixes (broken links, incorrect sphinx...
• #8837: DOC: Spell out note for hstack
• #8840: DOC: update docs and comments for move of mailing list to python.org
• #8843: MAINT: Use AxisError in more places
• #8844: DOC: Spell out note for dstack
• #8845: DOC: Add release note about np.real and np.conj
• #8846: BUG: Buttress handling of extreme values in randint
• #8847: DOC: Preliminary edit of 1.13.0 release notes.
• #8850: DOC: Updated doc of nonzero()
• #8852: MAINT: restore auto-vectorization of inplace operations
• #8854: MAINT: Remove manual expansion of template loop for some ufuncs
• #8857: DOC: remove empty jargon reference in glossary
• #8859: DOC: Fixed README formatting
• #8861: MAINT: Include the function name in all argument error messages
• #8862: BUG: do not memcpy ptr to freed object
• #8870: TST: Respect compiler customizations
• #8871: DOC: Replace line that was errantly removed in #8850
• #8873: BUG: Make runtests.py --shell behave better on windows
• #8874: TST: Use explicit NaT in test_structure_format
• #8876: MAINT: Minor ufunc cleanup
• #8883: BUG: Ensure Errors are correctly checked when PyFloat_AsDouble...
• #8884: BUG: Check for errors when PyInt_AsLong is called in np.random
• #8885: ENH: add support for python3.6 memory tracing
• #8886: ENH: add np.block to improve upon np.bmat
• #8888: BUG: Don't modify types after PyType_Ready
• #8890: DOC: proposed fixes for issues #7622 and #7914
• #8894: MAINT: Use PyArray_FROM* macros
• #8895: BUG: return values of exec_command were swapped
• #8896: ENH: do integer**2. inplace
• #8897: ENH: don't rebuild unchanged files
• #8898: BUG: Move ctypes ImportError catching to appropriate place
• #8900: Fix typos.
• #8903: BUG: Fix setitem on UNICODE, STRING, and LONGDOUBLE
• #8905: BUG: Correctly distinguish between 0d arrays and scalars in MaskedArray.__getitem__
• #8907: COMPAT: notify garbage collector when memory is allocated
• #8911: BUG: check_api_dict does not correctly handle tuple values
• #8914: DOC: Replace reference to np.swapaxis with np.swapaxes
• #8918: DEP: deprecate calling ma.argsort without an axis
• #8919: MAINT, TST: Remove duplicated code for testing the two types...
• #8921: MAINT: avoid memcpy when i == j
• #8925: DOC: Fix incorrect call to set_printoptions
• #8928: BUG: runtests --bench fails on windows
• #8929: BENCH: Masked array benchmarks
• #8939: DEP: Deprecate np.ma.MaskedArray.mini
• #8942: DOC: stop referring to 'S' dtype as string
• #8948: DEP: Deprecate NPY_CHAR
• #8949: REL: add python_requires to setup.py
• #8951: ENH: Add ufunc.identity for hypot and logical_xor
• #8953: DEP: Add back ndarray.__[sg]etslice__, but deprecate it
• #8959: DEP: Remove alter/restore dot methods
• #8961: MAINT: Update Intel compiler options.
• #8962: DOC: Wrong return type of np.random.choice and wrong variable...
• #8963: BUG: Prevent crash on repr of recursive array
• #8964: BUG: don't create array with invalid memory in where
• #8967: ENH: add np.positive ufunc
• #8971: BUG: do not change size 0 description when viewing data
• #8976: BUG: Prevent VOID_copyswapn ignoring strides
• #8978: TST: enable shadowed test
• #8980: DOC: Correct shape of edges in np.histogram2d
• #8988: DOC: Explain the behavior of diff on unsigned types
• #8989: ENH: Print object arrays containing lists unambiguously
• #8996: BUG/DEP: Make ufunclike functions more ufunc-like
• #8997: TST: fix io test that doesn't close file
• #8998: DOC: Use ` instead of * to refer to a function parameter.
• #8999: TST: Enable NPY_RELAXED_STRIDES_DEBUG environment variable.
• #9002: MAINT: Document ufunc(where=...) as defaulting to True
• #9012: MAINT: Set the __name__ of generated methods
• #9013: BUG: Fix np.lib.nanfunctions on object arrays
• #9014: BUG: __array_ufunc__= None -> TypeError
• #9015: ENH: Use __array_ufunc__ = None in polynomial convenience classes.
• #9021: BUG: Make ndarray inplace operators forward calls when needed.
• #9024: DOC: Correct default stop index value for negative stepping.
• #9026: ENH: Show full PEP 457 argument lists for ufuncs
• #9027: DOC: update binary-op / ufunc interactions and recommendations...
• #9038: BUG: check compiler flags to determine the need for a rebuild
• #9039: DOC: actually produce docs for as_strided
• #9050: BUG: distutils, add compatibility python parallelization
• #9054: BUG: Various fixes to _dtype_from_pep3118
• #9058: MAINT: Update FutureWarning message.
• #9060: DEP: deprecate ndarray.conjugate's no-op fall through for non-numeric...
• #9061: BUG: ndarray.conjugate broken for custom dtypes (unlike np.conjugate)
• #9062: STY: two blank lines between classes per PEP8
• #9063: ENH: add np.divmod ufunc
• #9070: BUG: Preserve field order in join_by, avoids FutureWarning
• #9072: BUG: if importing multiarray fails, don't discard the error message
• #9074: MAINT: Python 3.6 invalid escape sequence deprecation fixes
• #9075: ENH: Spelling fixes
• #9077: BUG: Prevent stackoverflow on self-containing arrays
• #9080: MAINT, DOC: Update 1.13.0 release notes and .mailmap
• #9087: BUG: __array_ufunc__ should always be looked up on the type,...
• #9091: MAINT: refine error message for __array_ufunc__ not implemented
• #9093: BUG remove memory leak in array ufunc override.
• #9097: TST: fix test_basic failure on Windows
• #9111: BUG: Array ufunc reduce out tuple
• #9123: DOC: update 1.13 release note for MaskedArray, masked constants...
• #9124: BUG: Do not elide complex abs() for 1.13
• #9129: BUG: ndarray.__pow__ does not check result of fast_scalar_power
• #9133: DEP: Deprecate incorrect behavior of expand_dims.
• #9135: BUG: delay calls of array repr in getlimits
• #9136: BUG: Compilation crashes in MSVC when LIB or INCLUDE is not set
• #9173: BUG: have as_strided() keep custom dtypes
• #9175: BUG: ensure structured ndarray.__eq__,__ne__ defer when appropriate.
• #9196: BUG: pull request 9087 modifies a tuple after use
• #9199: DOC: Update bincount docs to reflect gh-8348 (backport)