summaryrefslogtreecommitdiff
path: root/numpy
Commit message (Collapse)AuthorAgeFilesLines
* Merge pull request #12713 from mattip/gh-12711Tyler Reddy2019-01-103-19/+25
|\ | | | | BUG: loosen kwargs requirements in ediff1d
| * BUG: loosen kwargs requirements in ediff1dmattip2019-01-103-19/+25
| |
* | Merge pull request #12545 from bashtage/choice-nan-protectCharles Harris2019-01-102-1/+9
|\ \ | | | | | | BUG: Ensure probabilities are not NaN in choice
| * | BUG: Ensure probabilities are not NaN in choiceKevin Sheppard2018-12-142-1/+9
| | | | | | | | | | | | | | | | | | Add a check for NaN probabilities in random.choice closes #11250
* | | Merge pull request #12696 from seberg/bufferinfo_cache_dealloc_voidMatti Picus2019-01-104-0/+6
|\ \ \ | | | | | | | | BUG: Fix leak of void scalar buffer info
| * | | BUG: Followup fixes of leaks detected with valgrindSebastian Berg2019-01-092-0/+2
| | | | | | | | | | | | | | | | See also gh-12624
| * | | BUG: Free buffer shape information on error during creationSebastian Berg2019-01-081-0/+2
| | | | | | | | | | | | | | | | | | | | When the buffer was created and an error was found during the format string construction, the buffer was not freed correctly.
| * | | BUG: Fix leak of void scalar buffer infoSebastian Berg2019-01-081-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The vois scalar was missing the clearing logic for the buffer info cache (or storage). It should maybe be checked whether we can replace all of this logic when not supporting Python 2 anymore.
* | | | Merge pull request #12702 from rth/fft-3d-fortran-testCharles Harris2019-01-101-0/+38
|\ \ \ \ | |_|_|/ |/| | | TST: Check FFT results for C/Fortran ordered and non contigous input
| * | | SpellingRoman Yurchak2019-01-091-1/+1
| | | |
| * | | TST Check FFT for C/Fortran ordered and non contigous arraysRoman Yurchak2019-01-091-0/+38
| | | |
* | | | Merge pull request #12668 from anntzer/concat-error-messageMatti Picus2019-01-102-8/+27
|\ \ \ \ | | | | | | | | | | ENH: Improve error messages for non-matching shapes in concatenate.
| * | | | Improve error messages for non-matching shapes in concatenate.Antony Lee2019-01-092-8/+27
| | | | | | | | | | | | | | | | | | | | | | | | | ... by including the indices and shapes of the arrays whose shape differ.
* | | | | BUG: reference cycle in np.vectorize (#11977)Matti Picus2019-01-094-12/+94
| |/ / / |/| | | | | | | | | | | | | | | | | | | | | | | This implements cyclic support by adding `tp_traverse` to ufuns which may contain a user provided object function (`np.frompyfunc`). Ufuncs that do not add this are not added to the circular reference count tracking. The ufunc does not need to implement `tp_clear` because it is an immutable object.
* | | | Merge pull request #12679 from mattip/remove-warningsSebastian Berg2019-01-094-141/+68
|\ \ \ \ | | | | | | | | | | ENH: remove "Invalid value" warnings from median, percentile
| * | | | ENH: remove "Invalid value" warnings from median, percentilemattip2019-01-064-141/+68
| | | | |
* | | | | Merge pull request #12684 from mattip/benchmark-ufuncsSebastian Berg2019-01-092-3/+4
|\ \ \ \ \ | | | | | | | | | | | | DEV: remove _arg from public API, add matmul to benchmark ufuncs
| * | | | | DEV: remove _arg from public API, add matmul to benchmark ufuncsmattip2019-01-072-3/+4
| |/ / / /
* | | | | DOC: improve comments in copycast_isalignedAllan Haldane2019-01-081-1/+10
| | | | |
* | | | | Merge pull request #12685 from mreineck/add_pocketfftMatti Picus2019-01-082-5/+23
|\ \ \ \ \ | | | | | | | | | | | | BUG: Make pocketfft handle long doubles.
| * | | | | tweak testMartin Reinecke2019-01-081-5/+7
| | | | | |
| * | | | | fix #12663Martin Reinecke2019-01-082-5/+21
| | |_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tweak and add test better fix fix cleanup, additional test fix test
* | | | | Merge pull request #12677 from ahaldane/fix_more_uint_alnCharles Harris2019-01-084-20/+73
|\ \ \ \ \ | |/ / / / |/| | | | MAINT: Further fixups to uint alignment checks
| * | | | MAINT: Further fixups to uint alignment checksAllan Haldane2019-01-064-20/+73
| | |/ / | |/| |
* | | | Merge pull request #12680 from eric-wieser/fix-PyArray_RegisterCanCast-mallocMatti Picus2019-01-071-8/+25
|\ \ \ \ | |_|/ / |/| | | BUG: Ensure failing memory allocations are reported
| * | | BUG: Ensure failing memory allocations are reportedEric Wieser2019-01-061-8/+25
| | | | | | | | | | | | | | | | Despite the strange tone, gh-12674 did identify a series of low-impact bugs in PyArray_RegisterCanCast
* | | | Merge pull request #12672 from charris/fix-selected_real_kind-for-sparcCharles Harris2019-01-051-1/+1
|\ \ \ \ | | | | | | | | | | BUG: Add 'sparc' to platforms implementing 16 byte reals.
| * | | | BUG: Add 'sparc' to platforms implementing 16 byte reals.Charles Harris2019-01-051-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | This is for the _selected_real_kind_func function in f2py/crackfortran.py
* | | | | TST: Fix endianness in unstuctured_to_structured testAllan Haldane2019-01-051-3/+3
| |_|/ / |/| | |
* | | | Merge pull request #12239 from daten-kieker/polyval_2477Marten van Kerkwijk2019-01-042-2/+17
|\ \ \ \ | | | | | | | | | | BUG: polyval returned non-masked arrays for masked input.
| * | | | BUG: polyval returned Non-Masked Arrays for Masked Input.Joachim Hereth2018-10-212-2/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fix will preserve subtypes of ndarray when given as input (x) to the polyval function. In particular, the results for masked values of a masked array will be masked. Fixes #2477.
* | | | | ENH: allow where in reduce operations.Marten van Kerkwijk2019-01-038-89/+241
| |/ / / |/| | | | | | | | | | | | | | | In this implementation, if the ufunc does not have an identity, it needs an initial vavlue to be supplied.
* | | | Merge pull request #12626 from ahaldane/further_uint_align_fixCharles Harris2019-01-039-46/+99
|\ \ \ \ | | | | | | | | | | BUG: fix uint alignment asserts in lowlevel loops
| * | | | BUG: fix uint alignment asserts in lowlevel loopsAllan Haldane2019-01-039-46/+99
| | | | |
* | | | | BUG: Fix incorrect/missing reference cleanups found using valgrind (#12624)Sebastian Berg2019-01-0318-30/+83
| | | | | | | | | | | | | | | squash-merged
* | | | | DOC, TST: clean up matplotlib importsTyler Reddy2019-01-025-25/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * as requested by review in gh-12634, the vast majority of docstring matplotlib imports can be simplified to a single line
* | | | | DOC, TST: remove agg setting from docsTyler Reddy2019-01-024-10/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * there is no longer any usage of "agg" backend switching in our docstrings because this backend is already activated in the refguide_check machinery
* | | | | ENH: add "max difference" messages to np.testing.assert_array_equal (#12591)Stephan Hoyer2019-01-013-135/+185
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Example behavior: >>> x = np.array([1, 2, 3]) >>> y = np.array([1, 2, 3.0001]) >>> np.testing.assert_allclose(x, y) AssertionError: Not equal to tolerance rtol=1e-07, atol=0 Mismatch: 33.3% Max absolute difference: 0.0001 Max relative difference: 3.33322223e-05 x: array([1, 2, 3]) y: array([1. , 2. , 3.0001]) Motivation: when writing numerical algorithms, I frequently find myself experimenting to pick the right value of `atol` and `rtol` for `np.testing.assert_allclose()`. If I make the tolerance too generous, I risk missing regressions in accuracy, so I usually try to pick the smallest values for which tests pass. This change immediately reveals appropriate values to use for these parameters, so I don't need to guess and check.
* | | | | Merge pull request #12308 from tylerjereddy/timedelta64_floordivMatti Picus2019-01-015-3/+118
|\ \ \ \ \ | | | | | | | | | | | | ENH: add mm->q floordiv
| * | | | | ENH: add mm->q floordivTyler Reddy2018-12-075-3/+118
| | |_|/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | * add support for floor division between timedelta64 (m8) operands with generic or specific units; type signature is mm->q
* | | | | Merge pull request #12551 from rgommers/append-warnMatti Picus2018-12-312-4/+52
|\ \ \ \ \ | |_|/ / / |/| | | | MAINT: add warning to numpy.distutils for LDFLAGS append behavior.
| * | | | TST: suppress test warnings for appending flag behavior better.Ralf Gommers2018-12-271-3/+8
| | | | |
| * | | | MAINT: add warning to numpy.distutils for LDFLAGS append behavior.Ralf Gommers2018-12-272-4/+47
| | | | | | | | | | | | | | | | | | | | See gh-7427
* | | | | Merge pull request #12618 from ahaldane/test_12byte_alignmentCharles Harris2018-12-271-2/+2
|\ \ \ \ \ | | | | | | | | | | | | MAINT: add test for 12-byte alignment
| * | | | | MAINT: add test for 12-byte alignmentAllan Haldane2018-12-261-2/+2
| | | | | |
* | | | | | Merge pull request #12601 from abdurrehman11/numpy-delete-docsAllan Haldane2018-12-271-1/+1
|\ \ \ \ \ \ | |_|/ / / / |/| | | | | DOC: Correct documentation of `numpy.delete` obj parameter.
| * | | | | Refactored numpy.delete function docsAbdur Rehman2018-12-211-1/+1
| | | | | |
* | | | | | BUG: Add missing free in ufunc deallocSebastian Berg2018-12-261-0/+2
| |/ / / / |/| | | |
* | | | | Merge pull request #12611 from mattip/align-longdoubleCharles Harris2018-12-262-3/+9
|\ \ \ \ \ | | | | | | | | | | | | BUG: longdouble with elsize 12 is never uint alignable
| * | | | | MAINT: improve commentsmattip2018-12-262-3/+4
| | | | | |
View Lorry Controller Status