| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | MAINT: Do not forward `__(deep)copy__` calls of `_GenericAlias` to the ↵ | Bas van Beek | 2021-11-10 | 2 | -0/+18 |
| | | | | | | | wrapped type Adapt to the python 3.9.8 changes made in bpo-45167. | ||||
| * | Merge pull request #20343 from mattip/alloc-version | Sebastian Berg | 2021-11-10 | 7 | -10/+73 |
| |\ | | | | | ENH: add a 'version' field to PyDataMem_Handler | ||||
| | * | ENH: add a 'version' field to PyDataMem_Handler | mattip | 2021-11-10 | 7 | -10/+73 |
| | | | |||||
| * | | Merge pull request #20331 from bzah/fix/quantile-default-lerp-method | Sebastian Berg | 2021-11-09 | 2 | -5/+15 |
| |\ \ | |/ |/| | MAINT: Update quantile default lerp method | ||||
| | * | TST: Add test for max ulp in default quantile calculation | abel | 2021-11-09 | 1 | -0/+8 |
| | | | |||||
| | * | MTH: Update quantile default lerp method | abel | 2021-11-09 | 1 | -5/+7 |
| | | | | | | | | | | | | | For method 7 of H&F, using `(n - 1) * quantiles` instead of the usual method gives a more accurate result. | ||||
| * | | Merge pull request #20333 from lazka/mingw-remove-old-version-checks | Matti Picus | 2021-11-09 | 1 | -69/+12 |
| |\ \ | | | | | | | DEP: remove code for supporting GCC <4 in Mingw32CCompiler | ||||
| | * | | DEP: remove code for supporting GCC <4 in Mingw32CCompiler | Christoph Reiter | 2021-11-09 | 1 | -69/+12 |
| | |/ | | | | | | | | | | | | | The last GCC 3.x release before 4.0 was 3.4 which was released 2004, so 17 years ago. Removes all code only used with those old toolchains. | ||||
| * | | Merge pull request #19083 from hameerabbasi/dlpack | Sebastian Berg | 2021-11-09 | 14 | -17/+820 |
| |\ \ | |/ |/| | ENH: Implement the DLPack Array API protocols for ndarray. | ||||
| | * | MAINT: Simplify `byte_offset` handling in dlpack.h and add comment | Sebastian Berg | 2021-11-09 | 1 | -24/+16 |
| | | | | | | | | | | | | | | | If `byte_offset = 0` is forced anyway, there is no point in trying to preserve a previous `data` information from the capsule. (And probably it should have used a base array also, and not just a base DLPack capsule, anyway.) | ||||
| | * | add release note, error out if offset is used | mattip | 2021-11-09 | 1 | -1/+7 |
| | | | |||||
| | * | make a.__dlpack__() fail if a is readonly | mattip | 2021-11-04 | 2 | -1/+13 |
| | | | |||||
| | * | change from_dlpack to _dlpack, remove unused header | mattip | 2021-11-02 | 11 | -49/+32 |
| | | | |||||
| | * | Updates | Matti Picus | 2021-11-02 | 2 | -5/+3 |
| | | | | | | | | | | | Co-authored-by: Sebastian Berg <sebastian@sipsolutions.net> Co-authored-by: Bas van Beek <43369155+BvB93@users.noreply.github.com> | ||||
| | * | BUG: fixes from review | mattip | 2021-11-02 | 1 | -40/+64 |
| | | | |||||
| | * | move dlpack functions to a new file | mattip | 2021-11-02 | 7 | -393/+419 |
| | | | |||||
| | * | BUG: Offset not properly stored/computed. | Hameer Abbasi | 2021-11-02 | 2 | -10/+29 |
| | | | |||||
| | * | MAINT: Robustify dlpack_capsule_deleter and add comments. | Hameer Abbasi | 2021-11-02 | 2 | -8/+34 |
| | | | |||||
| | * | BUG, TST: Device bugfix and test __dl_device__. | Hameer Abbasi | 2021-11-02 | 3 | -14/+32 |
| | | | |||||
| | * | BUG: Fix handling of C-contiguous and 1-element arrays. | Hameer Abbasi | 2021-11-02 | 3 | -10/+18 |
| | | | |||||
| | * | TST: Improve testing coverage for DLPack. | Hameer Abbasi | 2021-11-02 | 1 | -0/+11 |
| | | | |||||
| | * | MAINT: Move around code so that there are no more unused warnings. | Hameer Abbasi | 2021-11-02 | 3 | -25/+24 |
| | | | |||||
| | * | MAINT: Split up capsule deleter. | Hameer Abbasi | 2021-11-02 | 2 | -7/+19 |
| | | | |||||
| | * | MAINT: Add URL to DLPack GitHub. | Hameer Abbasi | 2021-11-02 | 1 | -0/+2 |
| | | | |||||
| | * | MAINT, BUG: Documentation for DLPack protocol and refcounting bug fixes. | Hameer Abbasi | 2021-11-02 | 6 | -25/+58 |
| | | | |||||
| | * | ENH, TST: Add the from_dlpack method and test DLPack. | Hameer Abbasi | 2021-11-02 | 8 | -14/+287 |
| | | | |||||
| | * | ENH: Add the __dlpack__ and __dlpack_device__ methods to ndarray. | Hameer Abbasi | 2021-11-02 | 3 | -0/+361 |
| | | | |||||
| * | | Merge pull request #19429 from defoishugo/fix_multiarray_leaks | Matti Picus | 2021-11-09 | 3 | -5/+12 |
| |\ \ | | | | | | | BUG: Fix some multiarray leaks | ||||
| | * | | Fixups based on review comments. | Sebastian Berg | 2021-09-29 | 3 | -8/+11 |
| | | | | |||||
| | * | | BUG: fix a PyArray_CastScalarToCtype leak | Hugo Defois | 2021-09-29 | 1 | -0/+2 |
| | | | | |||||
| | * | | BUG: fix a PyInit__multiarray_umath leak | Hugo Defois | 2021-09-29 | 1 | -0/+1 |
| | | | | |||||
| | * | | BUG: fix a PyArray_DiscoverDTypeAndShape_Recursive leak | Hugo Defois | 2021-09-29 | 1 | -0/+1 |
| | | | | |||||
| * | | | DOC: Remove non-existent alias | abel | 2021-11-08 | 3 | -13/+10 |
| | | | | |||||
| * | | | MAINT: Remove Python <3.8 support from C | Sebastian Berg | 2021-11-07 | 4 | -88/+3 |
| | | | | | | | | | | | | | | | | | | | Remove version switches for support of Python versions before 3.8. The biggest deletion is that we can use Pythons `PyVectorcall_Call` to support the old `*args, **kwargs` style call convention in ufuncs. | ||||
| * | | | Merge pull request #20292 from seiko2plus/replace_raw_ceil | Matti Picus | 2021-11-07 | 5 | -33/+26 |
| |\ \ \ | | | | | | | | | SIMD: replace raw SIMD of ceil with universal intrinsics | ||||
| | * | | | SIMD: replace raw SIMD of ceil with universal intrinsics | Sayed Adel | 2021-11-06 | 5 | -33/+26 |
| | | | | | |||||
| * | | | | Merge pull request #20308 from BvB93/is_pyston | Charles Harris | 2021-11-05 | 3 | -3/+3 |
| |\ \ \ \ | | | | | | | | | | | MAINT: Add `IS_PYSTON` to `np.testing.__all__` | ||||
| | * | | | | MAINT: Add missing annotations for `IS_PYSTON` | Bas van Beek | 2021-11-05 | 1 | -0/+1 |
| | | | | | | |||||
| | * | | | | MAINT: Add `IS_PYSTON` to `np.testing.__all__` | Bas van Beek | 2021-11-05 | 2 | -3/+2 |
| | | | | | | |||||
| * | | | | | Merge pull request #18203 from anntzer/fq | Sebastian Berg | 2021-11-05 | 1 | -1/+2 |
| |\ \ \ \ \ | | | | | | | | | | | | | MAINT: Speedup np.quantile. | ||||
| | * | | | | | PERF: Speedup np.quantile. | Antony Lee | 2021-11-05 | 1 | -1/+2 |
| | |/ / / / | | | | | | | | | | | | | | | | | | | | | | | | | | Avoiding the unnecessary calls to moveaxis() speed up `np.quantile(x, .5)` (`x = np.random.rand(1000)`) by ~10% (although there's a lot of variability) for me. | ||||
| * | | | | | MAINT: Add annotations for a missing `percentile` interpolation kind: ↵ | Bas van Beek | 2021-11-05 | 1 | -0/+1 |
| |/ / / / | | | | | | | | | | | | | `"inclusive"` | ||||
| * | | | | Merge pull request #19857 from bzah/feature/rewrite_percentile | Sebastian Berg | 2021-11-04 | 4 | -240/+842 |
| |\ \ \ \ | | | | | | | | | | | MAINT, ENH: Refactor percentile and quantile methods | ||||
| | * | | | | DOC: Refer to the quantile/percentile notes for nan versions | Sebastian Berg | 2021-11-04 | 2 | -201/+4 |
| | | | | | | |||||
| | * | | | | MAINT: Remove unnecessary import of _QuantileInterpolation | Sebastian Berg | 2021-11-04 | 1 | -1/+0 |
| | | | | | | |||||
| | * | | | | TST: Make use of clongdouble and longdouble | abel | 2021-11-04 | 1 | -2/+2 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | On some platforms float128 and complex256 do not exist. Using (c)longdouble aliases should work on all platforms. | ||||
| | * | | | | DOC: Update _InterpolationKind | abel | 2021-11-04 | 1 | -1/+9 |
| | | | | | | |||||
| | * | | | | TST: Add parametrize for interpolation methods | abel | 2021-11-04 | 1 | -91/+43 |
| | | | | | | |||||
| | * | | | | TST: Add extrapolation tests | abel | 2021-11-04 | 1 | -0/+9 |
| | | | | | | |||||
| | * | | | | MAINT: Clean following PR comments | abel | 2021-11-04 | 2 | -17/+7 |
| | | | | | | |||||
 |
index : delta/python-packages/numpy.git | |
| github.com: numpy/numpy.git |
| summaryrefslogtreecommitdiff |