History for numpy/_core/code_generators - numpy/numpy · GitHub

github.com

Tin mới

Merge pull request #32475 from AltuRome/add-boolean-funcationality-to-positive-Ufunc

Merge pull request #32475 from AltuRome/add-boolean-funcationality-to-positive-Ufunc

Update numpy/_core/code_generators/generate_umath.py

Update numpy/_core/code_generators/generate_umath.py

BUG/TST: np.positive now registers boolean objects. Closes #32100

BUG/TST: np.positive now registers boolean objects. Closes #32100

ENH,SIMD: Use Google Highway for timedelta division SIMD

ENH,SIMD: Use Google Highway for timedelta division SIMD

ENH,SIMD: Replace libdivide timedelta division with dispatched SIMD

ENH,SIMD: Replace libdivide timedelta division with dispatched SIMD

ENH: implement `np.unwrap` as a generalized ufunc (#31848)

ENH: implement `np.unwrap` as a generalized ufunc (#31848)

BUG: fix numpy datetime cython APIs to be compatible with older versions

BUG: fix numpy datetime cython APIs to be compatible with older versions

BLD: fix build failure with PYTHONSAFEPATH=1 (#30731)

BLD: fix build failure with PYTHONSAFEPATH=1 (#30731)

BEG: Begin NumPy 2.6.0 development (#31492)

BEG: Begin NumPy 2.6.0 development (#31492)

ENH: Create a path for legacy dtypes to transition to new DType API (#31302)

ENH: Create a path for legacy dtypes to transition to new DType API (#31302)

ENH: Implement opaque PyObject ABI support (#31091)

ENH: Implement opaque PyObject ABI support (#31091)

BUG: raise warning when `np.timedelta64()` is called (#31213)

BUG: raise warning when `np.timedelta64()` is called (#31213)

ENH: mark ``_import_array`` function as inline (#31258)

ENH: mark ``_import_array`` function as inline (#31258)

BUG: Remove thread-local allocator cache implementation (#31237)

BUG: Remove thread-local allocator cache implementation (#31237)

BUG: Use dict to store ufunc loops (fixes thread-safety) (#31184)

BUG: Use dict to store ufunc loops (fixes thread-safety) (#31184)

Merge branch 'main' into fix-numpy/_core-UP031

Merge branch 'main' into fix-numpy/_core-UP031

ENH: Improve performance of ufuncs using NPY_METH_NO_FLOATINGPOINT_ERRORS (#30987)

ENH: Improve performance of ufuncs using NPY_METH_NO_FLOATINGPOINT_ERRORS (#30987)

ENH: Add internal real/imag ufuncs, use for attributes and refine object dtype (#30984)

ENH: Add internal real/imag ufuncs, use for attributes and refine object dtype (#30984)

MAINT: fix two minor issues noticed when touching the C API setup

MAINT: fix two minor issues noticed when touching the C API setup

DOC: Correct typos in code comments [skip azp][skip cirrus][skip actions]

DOC: Correct typos in code comments [skip azp][skip cirrus][skip actions]

MAINT: Replace %-formatting with f-strings in numpy/_core (UP031)

MAINT: Replace %-formatting with f-strings in numpy/_core (UP031)

Merge pull request #30194 from WarrenWeckesser/sign-timedelta64

Merge pull request #30194 from WarrenWeckesser/sign-timedelta64

ENH: enable alloc cache on free-threading builds (#30499)

ENH: enable alloc cache on free-threading builds (#30499)

MAINT: Enable linting with ruff E501

MAINT: Enable linting with ruff E501

DOC: prefer passing scalar types to ``dtype=`` in examples (#30394)

DOC: Clarify `np.negative`

BUG: Change return type of sign(timedelta64) to float64.

DOC: make descriptions of hyperbolic functions consistent (#29297)

remove numpy.distutils

BEG, MAINT: Begin NumPy 2.5 development (#30333)

MAINT: Migrate einsum.c.src to C++ (einsum.cpp) (#30142)

BUG: Avoid compilation error of wrapper file generated with SWIG >= 4.4 (#30128)

ENH: Add registration for sort and argsort in new convenience ufunc registration function

ENH: add a casting option 'same_value' and use it in np.astype (#29129)

DOC: Add 'See Also' refs for sign, copysign and signbit. (#29489)