21 Pull requests merged by 8 people

• TYP: Fix inconsistent NDArray[float64].__[r]truediv__ return type with float

  #28706 merged Apr 14, 2025

• TYP: Fix false rejection of NDArray[object_].__abs__()

  #28705 merged Apr 14, 2025

• TYP: Fix false rejection of NDArray[object_].__abs__()

  #28699 merged Apr 14, 2025

• TYP: Fix inconsistent NDArray[float64].__[r]truediv__ return type with float

  #28702 merged Apr 14, 2025

• TYP: replace _SCT with _ScalarT

  #28690 merged Apr 13, 2025

• DOC: linalg.matrix_transpose: add alias note

  #28695 merged Apr 12, 2025

• BLD: fix meson_version warning

  #28693 merged Apr 12, 2025

• TYP: replace _DType with _DTypeT

  #28684 merged Apr 11, 2025

• TYP: rename _ShapeType TypeVar to _ShapeT

  #28688 merged Apr 11, 2025

• TYP: Type MaskedArray.take and np.ma.take

  #28680 merged Apr 11, 2025

• TYP: add missing "b1" literals for dtype[bool]

  #28683 merged Apr 10, 2025

• SIMD: Resolve Highway QSort symbol linking error on aarch32/ASIMD

  #28682 merged Apr 10, 2025

• TYP: Type MaskedArray.partition and MaskedArray.argpartition

  #28677 merged Apr 10, 2025

• TYP: fix mypy test failures

  #28676 merged Apr 9, 2025

• TYP: add missing "b1" literals for dtype[bool]

  #28674 merged Apr 9, 2025

• TYP: default to dtype[Any]

  #28669 merged Apr 9, 2025

• TYP: replace _ArrayType with _ArrayT

  #28668 merged Apr 9, 2025

• ENH: Improve np.linalg.det performance

  #28649 merged Apr 9, 2025

• SIMD: Fix Highway QSort symbol linking error on aarch32/ASIMD

  #28671 merged Apr 9, 2025

• MAINT: Bump github/codeql-action from 3.28.14 to 3.28.15

  #28672 merged Apr 8, 2025

• TYP: replace _ScalarType with _SCT

  #28667 merged Apr 8, 2025

6 Pull requests opened by 6 people

• TYP: fix and improve ``numpy.lib._type_check_impl``

  #28675 opened Apr 9, 2025

• MAINT: Deprecate `.T` property for non-2dim arrays and scalars

  #28678 opened Apr 10, 2025

• ENH: Improve performance of np.linalg._linalg._commonType

  #28686 opened Apr 10, 2025

• TYP: Type ``MaskedArray.__ge__``

  #28689 opened Apr 11, 2025

• BUG: Respect print options in masked array formatting

  #28698 opened Apr 13, 2025

• MAINT: Improve float16 and float32 printing

  #28703 opened Apr 14, 2025

5 Issues closed by 5 people

• TYP: Loosing `np.float64` in numpy array type after division with a Python float

  #28701 closed Apr 14, 2025

• ENH: Allow disabling balancing in `linalg.eig` (Trac #1463)

  #2060 closed Apr 12, 2025

• ENH: Add `np.matrix_transpose`

  #24035 closed Apr 12, 2025

• BUG: Data type is printed when reordering the array using a list

  #28692 closed Apr 12, 2025

• Import error

  #28691 closed Apr 12, 2025

11 Issues opened by 11 people

• TYP: false positive when using `__ge__` between `NDArray[np.str_]`

  #28708 opened Apr 14, 2025

• BUG: `linalg.pinv` behaves differently on NumPy version 1.x and 2.x

  #28707 opened Apr 14, 2025

• BLD: Error installing Numpy 2.2.2 / 2.2.4 (2.2.x) on Raspberry Pi with Python3.13

  #28704 opened Apr 14, 2025

• BUG: f2py treats f90 module strings as variable names for dependent variables

  #28700 opened Apr 14, 2025

• DOC: docs build warnings

  #28694 opened Apr 12, 2025

• BUG: wrong errors e.g. "divide by zero encountered in matmul" on MacOS M4

  #28687 opened Apr 10, 2025

• BUG: print options not respected for masked arrays.

  #28685 opened Apr 10, 2025

• BUG: temporary elision heuristics are broken on Python 3.14

  #28681 opened Apr 10, 2025

• BUG: Floating point printing / string conversion is incorrect

  #28679 opened Apr 10, 2025

• BUG: mixed 0D integer array indices with 1D boolean indices leads to boolean index not being treating as `idx.nonzero()` per advanced indexing rules

  #28673 opened Apr 8, 2025

• BUG: StringDType `.str` property is confusing

  #28670 opened Apr 8, 2025

30 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• DOC: fixes classes decorated with set_module not showing its source on numpy's documentation (#28629)

  #28645 commented on Apr 14, 2025 • 7 new comments

• ENH, SIMD: Initial implementation of Highway wrapper

  #28622 commented on Apr 12, 2025 • 2 new comments

• ENH: Use openmp on x86-simd-sort to speed up np.sort and np.argsort

  #28619 commented on Apr 14, 2025 • 2 new comments

• MNT: add linter for thread-unsafe C API uses

  #28634 commented on Apr 13, 2025 • 1 new comment

• MAINT: Remove distutils CPU dispatcher compatibility code

  #28662 commented on Apr 13, 2025 • 0 new comments

• BUG: Fix npyv_signbit dispatch

  #28659 commented on Apr 9, 2025 • 0 new comments

• BLD: Fix ASIMD feature build on AArch32

  #28635 commented on Apr 10, 2025 • 0 new comments

• BUG: fix ``np.vectorize`` for object dtype

  #28624 commented on Apr 14, 2025 • 0 new comments

• ENH, API: New sorting mechanism for DType API

  #28516 commented on Apr 12, 2025 • 0 new comments

• ENH: Convert arithmetic from C universal intrinsics to C++ using Highway

  #27402 commented on Apr 14, 2025 • 0 new comments

• ENH: Add ``where`` argument to reduction functions that are missing it

  #26336 commented on Apr 14, 2025 • 0 new comments

• np.unique: support string dtypes

  #28364 commented on Apr 14, 2025 • 0 new comments

• ENH: implement is_integer() for np.ndarray

  #25766 commented on Apr 14, 2025 • 0 new comments

• BUG: Constructing `dtype=object` makes useless, slow calls to `__array__`

  #28651 commented on Apr 14, 2025 • 0 new comments

• Ecosystem compatibility with numpy 2.0

  #26191 commented on Apr 13, 2025 • 0 new comments

• ENH: Allow disabling error for `numpy.linalg.inv` when one matrix is singular in a stack

  #27035 commented on Apr 13, 2025 • 0 new comments

• DOC: any class decorated with `@set_module` does not show its source on automatically generated documentation

  #28629 commented on Apr 12, 2025 • 0 new comments

• BUG: MacOS matmul FPE heisenbug seen on CI

  #28227 commented on Apr 10, 2025 • 0 new comments

• TYP: `np.argmin` overloads could be more precise

  #28641 commented on Apr 10, 2025 • 0 new comments

• BUG: np.isclose with datetime64 and timedelta64 types fail with error messages that reveal implementation details

  #23157 commented on Apr 9, 2025 • 0 new comments

• BUG: `np.datetime64("nat", "ms")` does not use unit in `repr` (unlike timedelta)

  #28496 commented on Apr 9, 2025 • 0 new comments

• BUG: timedelta64 comparison is not transitive

  #28287 commented on Apr 9, 2025 • 0 new comments

• BUG: `timedelta64` with ns units converts to python `int` rather than a python `timedelta`

  #28596 commented on Apr 9, 2025 • 0 new comments

• TSK: Follow-up things for stringdtype

  #25693 commented on Apr 9, 2025 • 0 new comments

• BUG: Compilation failure on macOS with Emscripten

  #28642 commented on Apr 9, 2025 • 0 new comments

• DOC: SIMD information in ufunc documentation

  #16459 commented on Apr 9, 2025 • 0 new comments

• numpy.sum not stable enough sometimes (Kahan, math.fsum)

  #8786 commented on Apr 8, 2025 • 0 new comments

• ENH: Update numpy.typing.ArrayLike to allow specification of both dtype and shape/ndim

  #23745 commented on Apr 8, 2025 • 0 new comments

• Consider using pytest-mypy-plugins

  #16549 commented on Apr 8, 2025 • 0 new comments

• Add basic dtype typing support to the main numpy namespace

  #19252 commented on Apr 8, 2025 • 0 new comments