NumPy 2.0.0 Release Notes

NumPy 2.0.0 is the first major release since 2006. It is the result of
11 months of development since the last feature release and is the work
of 212 contributors spread over 1078 pull requests. It contains a large
number of exciting new features as well as changes to both the Python
and C APIs.

This major release includes breaking changes that could not happen in a
regular minor (feature) release - including an ABI break, changes to
type promotion rules, and API changes which may not have been emitting
deprecation warnings in 1.26.x. Key documents related to how to adapt to
changes in NumPy 2.0, in addition to these release notes, include:

The numpy-2-migration-guide
The Numpy 2.0-specific advice in for dpwmstream package authors

Highlights

Highlights of this release include:

New features:
- A new variable-length string dtype, numpy.dtypes.StringDType and a new
  numpy.strings namespace with performant ufuncs for string operations,
- Support for float32 and longdouble in all
  numpy.fft functions,
- Support for the array API standard in the main numpy
  namespace.
Performance improvements:
- Sorting functions sort, argsort,
  partition, argpartition have been
  accelerated through the use of the Intel x86-simd-sort and
  Google Highway libraries, and may see large (hardware-specific)
  speedups,
- macOS Accelerate support and binary wheels for macOS >=14, with
  significant performance improvements for linear algebra
  operations on macOS, and wheels that are about 3 times smaller,
- numpy.char fixed-length string operations have
  been accelerated by implementing ufuncs that also support
  numpy.dtypes.StringDType in addition to the
  fixed-length string dtypes,
- A new tracing and introspection API,
  numpy.lib.introspect.opt_func_info, to determine
  which hardware-specific kernels are available and will be
  dispatched to.
- numpy.save now uses pickle protocol version 4 for saving
  arrays with object dtype, which allows for pickle objects larger
  than 4GB and improves saving speed by about 5% for large arrays.
Python API improvements:
- A clear split between public and private API, with a new module
  structure and each public function now available in a single place.
- Many removals of non-recommended functions and aliases. This
  should make it easier to learn and use NumPy. The number of
  objects in the main namespace decreased by ~10% and in
  numpy.lib by ~80%.
- Canonical dtype names and a new numpy.isdtype` introspection
  function,
C API improvements:
- A new public C API for creating custom dtypes,
- Many outdated functions and macros removed, and private
  internals hidden to ease future extensibility,
- New, easier to use, initialization functions: PyArray_ImportNumPyAPI
  and PyUFunc_ImportUFuncAPI.
Improved behavior:
- Improvements to type promotion behavior was changed by adopting NEP 50.
  This fixes many user surprises about promotions which previously often
  depended on data values of input arrays rather than only their dtypes.
  Please see the NEP and the numpy-2-migration-guide for details as this
  change can lead to changes in output dtypes and lower precision results
  for mixed-dtype operations.
- The default integer type on Windows is now int64 rather than
  int32, matching the behavior on other platforms,
- The maximum number of array dimensions is changed from 32 to 64
Documentation:
- The reference guide navigation was significantly improved, and
  there is now documentation on NumPy's
  module structure,
- The building from source documentation was completely rewritten,

Furthermore there are many changes to NumPy internals, including
continuing to migrate code from C to C++, that will make it easier to
improve and maintain NumPy in the future.

The "no free lunch" theorem dictates that there is a price to pay for
all these API and behavior improvements and better future extensibility.
This price is:

Backwards compatibility. There are a significant number of breaking
changes to both the Python and C APIs. In the majority of cases,
there are clear error messages that will inform the user how to
adapt their code. However, there are also changes in behavior for
which it was not possible to give such an error message - these
cases are all covered in the Deprecation and Compatibility sections
below, and in the numpy-2-migration-guide.
Note that there is a ruff mode to auto-fix many things in Python
code.
Breaking changes to the NumPy ABI. As a result, binaries of packages
that use the NumPy C API and were built against a NumPy 1.xx release
will not work with NumPy 2.0. On import, such packages will see an
ImportError with a message about binary incompatibility.
It is possible to build binaries against NumPy 2.0 that will work at
runtime with both NumPy 2.0 and 1.x. See numpy-2-abi-handling for more
details.
All downstream packages that depend on the NumPy ABI are advised
to do a new release built against NumPy 2.0 and verify that that
release works with both 2.0 and 1.26 - ideally in the period between
2.0.0rc1 (which will be ABI-stable) and the final 2.0.0 release to
avoid problems for their users.

The Python versions supported by this release are 3.9-3.12.

NumPy 2.0 Python API removals

np.geterrobj, np.seterrobj and the related ufunc keyword
argument extobj= have been removed. The preferred replacement for
all of these is using the context manager with np.errstate():.
(gh-23922)
np.cast has been removed. The literal replacement for
np.cast[dtype](arg) is np.asarray(arg, dtype=dtype).
np.source has been removed. The preferred replacement is
inspect.getsource.
np.lookfor has been removed.
(gh-24144)
numpy.who has been removed. As an alternative for the removed
functionality, one can use a variable explorer that is available in
IDEs such as Spyder or Jupyter Notebook.
(gh-24321)
Warnings and exceptions present in numpy.exceptions,
e.g, numpy.exceptions.ComplexWarning,
numpy.exceptions.VisibleDeprecationWarning, are no
longer exposed in the main namespace.
Multiple niche enums, expired members and functions have been
removed from the main namespace, such as: ERR_*, SHIFT_*,
np.fastCopyAndTranspose, np.kernel_version, np.numarray,
np.oldnumeric and np.set_numeric_ops.
(gh-24316)
Replaced from ... import * in the numpy/__init__.py with
explicit imports. As a result, these main namespace members got
removed: np.FLOATING_POINT_SUPPORT, np.FPE_*, np.NINF,
np.PINF, np.NZERO, np.PZERO, np.CLIP, np.WRAP, np.WRAP,
np.RAISE, np.BUFSIZE, np.UFUNC_BUFSIZE_DEFAULT,
np.UFUNC_PYVALS_NAME, np.ALLOW_THREADS, np.MAXDIMS,
np.MAY_SHARE_EXACT, np.MAY_SHARE_BOUNDS, add_newdoc,
np.add_docstring and np.add_newdoc_ufunc.
(gh-24357)
Alias np.float_ has been removed. Use np.float64 instead.
Alias np.complex_ has been removed. Use np.complex128 instead.
Alias np.longfloat has been removed. Use np.longdouble instead.
Alias np.singlecomplex has been removed. Use np.complex64
instead.
Alias np.cfloat has been removed. Use np.complex128 instead.
Alias np.longcomplex has been removed. Use np.clongdouble
instead.
Alias np.clongfloat has been removed. Use np.clongdouble
instead.
Alias np.string_ has been removed. Use np.bytes_ instead.
Alias np.unicode_ has been removed. Use np.str_ instead.
Alias np.Inf has been removed. Use np.inf instead.
Alias np.Infinity has been removed. Use np.inf instead.
Alias np.NaN has been removed. Use np.nan instead.
Alias np.infty has been removed. Use np.inf instead.
Alias np.mat has been removed. Use np.asmatrix instead.
np.issubclass_ has been removed. Use the issubclass builtin
instead.
np.asfarray has been removed. Use np.asarray with a proper dtype
instead.
np.set_string_function has been removed. Use np.set_printoptions
instead with a formatter for custom printing of NumPy objects.
np.tracemalloc_domain is now only available from np.lib.
np.recfromcsv and recfromtxt are now only available from
np.lib.npyio.
np.issctype, np.maximum_sctype, np.obj2sctype,
np.sctype2char, np.sctypes, np.issubsctype were all removed
from the main namespace without replacement, as they where niche
members.
Deprecated np.deprecate and np.deprecate_with_doc has been
removed from the main namespace. Use DeprecationWarning instead.
Deprecated np.safe_eval has been removed from the main namespace.
Use ast.literal_eval instead.
(gh-24376)
np.find_common_type has been removed. Use numpy.promote_types or
numpy.result_type instead. To achieve semantics for the
scalar_types argument, use numpy.result_type and pass 0,
0.0, or 0j as a Python scalar instead.
np.round_ has been removed. Use np.round instead.
np.nbytes has been removed. Use np.dtype(<dtype>).itemsize
instead.
(gh-24477)
np.compare_chararrays has been removed from the main namespace.
Use np.char.compare_chararrays instead.
The charrarray in the main namespace has been deprecated. It can
be imported without a deprecation warning from np.char.chararray
for now, but we are planning to fully deprecate and remove
chararray in the future.
np.format_parser has been removed from the main namespace. Use
np.rec.format_parser instead.
(gh-24587)
Support for seven data type string aliases has been removed from
np.dtype: int0, uint0, void0, object0, str0, bytes0
and bool8.
(gh-24807)
The experimental numpy.array_api submodule has been removed. Use
the main numpy namespace for regular usage instead, or the
separate array-api-strict package for the compliance testing use
case for which numpy.array_api was mostly used.
(gh-25911)

`__array_prepare__` is removed

UFuncs called __array_prepare__ before running computations for normal
ufunc calls (not generalized ufuncs, reductions, etc.). The function was
also called instead of __array_wrap__ on the results of some linear
algebra functions.

It is now removed. If you use it, migrate to __array_ufunc__ or rely
on __array_wrap__ which is called with a context in all cases,
although only after the result array is filled. In those code paths,
__array_wrap__ will now be passed a base class, rather than a subclass
array.

(gh-25105)

Deprecations

np.compat has been deprecated, as Python 2 is no longer supported.
numpy.int8 and similar classes will no longer support conversion
of out of bounds python integers to integer arrays. For example,
conversion of 255 to int8 will not return -1. numpy.iinfo(dtype)
can be used to check the machine limits for data types. For example,
np.iinfo(np.uint16) returns min = 0 and max = 65535.
np.array(value).astype(dtype) will give the desired result.
np.safe_eval has been deprecated. ast.literal_eval should be
used instead.
(gh-23830)
np.recfromcsv, np.recfromtxt, np.disp, np.get_array_wrap,
np.maximum_sctype, np.deprecate and np.deprecate_with_doc have
been deprecated.
(gh-24154)
np.trapz has been deprecated. Use np.trapezoid or a
scipy.integrate function instead.
np.in1d has been deprecated. Use np.isin instead.
Alias np.row_stack has been deprecated. Use np.vstack directly.
(gh-24445)
__array_wrap__ is now passed arr, context, return_scalar and
support for implementations not accepting all three are deprecated.
Its signature should be
__array_wrap__(self, arr, context=None, return_scalar=False)
(gh-25409)
Arrays of 2-dimensional vectors for np.cross have been deprecated.
Use arrays of 3-dimensional vectors instead.
(gh-24818)
np.dtype("a") alias for np.dtype(np.bytes_) was deprecated. Use
np.dtype("S") alias instead.
(gh-24854)
Use of keyword arguments x and y with functions
assert_array_equal and assert_array_almost_equal has been
deprecated. Pass the first two arguments as positional arguments
instead.
(gh-24978)

`numpy.fft` deprecations for n-D transforms with None values in arguments

Using fftn, ifftn, rfftn, irfftn, fft2, ifft2, rfft2 or
irfft2 with the s parameter set to a value that is not None and
the axes parameter set to None has been deprecated, in line with the
array API standard. To retain current behaviour, pass a sequence [0,
..., k-1] to axes for an array of dimension k.

Furthermore, passing an array to s which contains None values is
deprecated as the parameter is documented to accept a sequence of
integers in both the NumPy docs and the array API specification. To use
the default behaviour of the corresponding 1-D transform, pass the value
matching the default for its n parameter. To use the default behaviour
for every axis, the s argument can be omitted.

(gh-25495)

`np.linalg.lstsq` now defaults to a new `rcond` value

numpy.linalg.lstsq now uses the new rcond value of the
machine precision times max(M, N). Previously, the machine precision
was used but a FutureWarning was given to notify that this change will
happen eventually. That old behavior can still be achieved by passing
rcond=-1.

(gh-25721)

Expired deprecations

The np.core.umath_tests submodule has been removed from the public
API. (Deprecated in NumPy 1.15)
(gh-23809)
The PyDataMem_SetEventHook deprecation has expired and it is
removed. Use tracemalloc and the np.lib.tracemalloc_domain
domain. (Deprecated in NumPy 1.23)
(gh-23921)
The deprecation of set_numeric_ops and the C functions
PyArray_SetNumericOps and PyArray_GetNumericOps has been expired
and the functions removed. (Deprecated in NumPy 1.16)
(gh-23998)
The fasttake, fastclip, and fastputmask ArrFuncs deprecation
is now finalized.
The deprecated function fastCopyAndTranspose and its C counterpart
are now removed.
The deprecation of PyArray_ScalarFromObject is now finalized.
(gh-24312)
np.msort has been removed. For a replacement, np.sort(a, axis=0)
should be used instead.
(gh-24494)
np.dtype(("f8", 1) will now return a shape 1 subarray dtype rather
than a non-subarray one.
(gh-25761)
Assigning to the .data attribute of an ndarray is disallowed and
will raise.
np.binary_repr(a, width) will raise if width is too small.
Using NPY_CHAR in PyArray_DescrFromType() will raise, use
NPY_STRING NPY_UNICODE, or NPY_VSTRING instead.
(gh-25794)

Compatibility notes

`loadtxt` and `genfromtxt` default encoding changed

loadtxt and genfromtxt now both default to encoding=None which may
mainly modify how converters work. These will now be passed str
rather than bytes. Pass the encoding explicitly to always get the new
or old behavior. For genfromtxt the change also means that returned
values will now be unicode strings rather than bytes.

(gh-25158)

`f2py` compatibility notes

f2py will no longer accept ambiguous -m and .pyf CLI
combinations. When more than one .pyf file is passed, an error is
raised. When both -m and a .pyf is passed, a warning is emitted
and the -m provided name is ignored.
(gh-25181)
The f2py.compile() helper has been removed because it leaked
memory, has been marked as experimental for several years now, and
was implemented as a thin subprocess.run wrapper. It was also one
of the test bottlenecks. See
gh-25122 for the full
rationale. It also used several np.distutils features which are
too fragile to be ported to work with meson.
Users are urged to replace calls to f2py.compile with calls to
subprocess.run("python", "-m", "numpy.f2py",... instead, and to
use environment variables to interact with meson. Native
files are also an
option.
(gh-25193)

Minor changes in behavior of sorting functions

Due to algorithmic changes and use of SIMD code, sorting functions with
methods that aren't stable may return slightly different results in
2.0.0 compared to 1.26.x. This includes the default method of
numpy.argsort and numpy.argpartition.

Removed ambiguity when broadcasting in `np.solve`

The broadcasting rules for np.solve(a, b) were ambiguous when b had
1 fewer dimensions than a. This has been resolved in a
backward-incompatible way and is now compliant with the Array API. The
old behaviour can be reconstructed by using
np.solve(a, b[..., None])[..., 0].

(gh-25914)

Modified representation for `Polynomial`

The representation method for
numpy.polynomial.polynomial.Polynomial was updated to
include the domain in the representation. The plain text and latex
representations are now consistent. For example the output of
str(np.polynomial.Polynomial([1, 1], domain=[.1, .2])) used to be
1.0 + 1.0 x, but now is 1.0 + 1.0 (-3.0000000000000004 + 20.0 x).

(gh-21760)

C API changes

The PyArray_CGT, PyArray_CLT, PyArray_CGE, PyArray_CLE,
PyArray_CEQ, PyArray_CNE macros have been removed.
PyArray_MIN and PyArray_MAX have been moved from
ndarraytypes.h to npy_math.h.
(gh-24258)
A C API for working with numpy.dtypes.StringDType
arrays has been exposed. This includes functions for acquiring and
releasing mutexes which lock access to the string data, as well as
packing and unpacking UTF-8 bytestreams from array entries.
NPY_NTYPES has been renamed to NPY_NTYPES_LEGACY as it does not
include new NumPy built-in DTypes. In particular the new string
DType will likely not work correctly with code that handles legacy
DTypes.
(gh-25347)
The C-API now only exports the static inline function versions of
the array accessors (previously this depended on using "deprecated
API"). While we discourage it, the struct fields can still be used
directly.
(gh-25789)
NumPy now defines PyArray_Pack to set an individual memory address.
Unlike PyArray_SETITEM this function is equivalent to setting an
individual array item and does not require a NumPy array input.
(gh-25954)
The ->f slot has been removed from PyArray_Descr. If you use this slot,
replace accessing it with PyDataType_GetArrFuncs (see its documentation
and the numpy-2-migration-guide). In some cases using other functions
like PyArray_GETITEM may be an alternatives.
PyArray_GETITEM and PyArray_SETITEM now require the import of
the NumPy API table to be used and are no longer defined in
ndarraytypes.h.
(gh-25812)
Due to runtime dependencies, the definition for functionality
accessing the dtype flags was moved from numpy/ndarraytypes.h and
is only available after including numpy/ndarrayobject.h as it
requires import_array(). This includes PyDataType_FLAGCHK,
PyDataType_REFCHK and NPY_BEGIN_THREADS_DESCR.
The dtype flags on PyArray_Descr must now be accessed through the
PyDataType_FLAGS inline function to be compatible with both 1.x
and 2.x. This function is defined in npy_2_compat.h to allow
backporting. Most or all users should use PyDataType_FLAGCHK which
is available on 1.x and does not require backporting. Cython users
should use Cython 3. Otherwise access will go through Python unless
they use PyDataType_FLAGCHK instead.
(gh-25816)

Datetime functionality exposed in the C API and Cython bindings

The functions NpyDatetime_ConvertDatetime64ToDatetimeStruct,
NpyDatetime_ConvertDatetimeStructToDatetime64,
NpyDatetime_ConvertPyDateTimeToDatetimeStruct,
NpyDatetime_GetDatetimeISO8601StrLen,
NpyDatetime_MakeISO8601Datetime, and
NpyDatetime_ParseISO8601Datetime have been added to the C API to
facilitate converting between strings, Python datetimes, and NumPy
datetimes in external libraries.

(gh-21199)

Const correctness for the generalized ufunc C API

The NumPy C API's functions for constructing generalized ufuncs
(PyUFunc_FromFuncAndData, PyUFunc_FromFuncAndDataAndSignature,
PyUFunc_FromFuncAndDataAndSignatureAndIdentity) take types and
data arguments that are not modified by NumPy's internals. Like the
name and doc arguments, third-party Python extension modules are
likely to supply these arguments from static constants. The types and
data arguments are now const-correct: they are declared as
const char *types and void *const *data, respectively. C code should
not be affected, but C++ code may be.

(gh-23847)

Larger `NPY_MAXDIMS` and `NPY_MAXARGS`, `NPY_RAVEL_AXIS` introduced

NPY_MAXDIMS is now 64, you may want to review its use. This is usually
used in a stack allocation, where the increase should be safe. However,
we do encourage generally to remove any use of NPY_MAXDIMS and
NPY_MAXARGS to eventually allow removing the constraint completely.
For the conversion helper and C-API functions mirroring Python ones such as
take, NPY_MAXDIMS was used to mean axis=None. Such usage must be replaced
with NPY_RAVEL_AXIS. See also migration_maxdims.

(gh-25149)

`NPY_MAXARGS` not constant and `PyArrayMultiIterObject` size change

Since NPY_MAXARGS was increased, it is now a runtime constant and not
compile-time constant anymore. We expect almost no users to notice this.
But if used for stack allocations it now must be replaced with a custom
constant using NPY_MAXARGS as an additional runtime check.

The sizeof(PyArrayMultiIterObject) no longer includes the full size of
the object. We expect nobody to notice this change. It was necessary to
avoid issues with Cython.

(gh-25271)

Required changes for custom legacy user dtypes

In order to improve our DTypes it is unfortunately necessary to break
the ABI, which requires some changes for dtypes registered with
PyArray_RegisterDataType. Please see the documentation of
PyArray_RegisterDataType for how to adapt your code and achieve
compatibility with both 1.x and 2.x.

(gh-25792)

New Public DType API

The C implementation of the NEP 42 DType API is now public. While the
DType API has shipped in NumPy for a few versions, it was only usable in
sessions with a special environment variable set. It is now possible to
write custom DTypes outside of NumPy using the new DType API and the
normal import_array() mechanism for importing the numpy C API.

See dtype-api for more details about the API. As always with a new feature,
please report any bugs you run into implementing or using a new DType. It is
likely that downstream C code that works with dtypes will need to be updated to
work correctly with new DTypes.

(gh-25754)

New C-API import functions

We have now added PyArray_ImportNumPyAPI and PyUFunc_ImportUFuncAPI
as static inline functions to import the NumPy C-API tables. The new
functions have two advantages over import_array and import_ufunc:

They check whether the import was already performed and are
light-weight if not, allowing to add them judiciously (although this
is not preferable in most cases).
The old mechanisms were macros rather than functions which included
a return statement.

The PyArray_ImportNumPyAPI() function is included in npy_2_compat.h
for simpler backporting.

(gh-25866)

Structured dtype information access through functions

The dtype structures fields c_metadata, names, fields, and
subarray must now be accessed through new functions following the same
names, such as PyDataType_NAMES. Direct access of the fields is not
valid as they do not exist for all PyArray_Descr instances. The
metadata field is kept, but the macro version should also be
preferred.

(gh-25802)

Descriptor `elsize` and `alignment` access

Unless compiling only with NumPy 2 support, the elsize and aligment
fields must now be accessed via PyDataType_ELSIZE,
PyDataType_SET_ELSIZE, and PyDataType_ALIGNMENT. In cases where the
descriptor is attached to an array, we advise using PyArray_ITEMSIZE
as it exists on all NumPy versions. Please see
migration_c_descr for more information.

(gh-25943)

NumPy 2.0 C API removals

npy_interrupt.h and the corresponding macros like NPY_SIGINT_ON
have been removed. We recommend querying PyErr_CheckSignals() or
PyOS_InterruptOccurred() periodically (these do currently require
holding the GIL though).
The noprefix.h header has been removed. Replace missing symbols
with their prefixed counterparts (usually an added NPY_ or
npy_).
(gh-23919)
PyUFunc_GetPyVals, PyUFunc_handlefperr, and PyUFunc_checkfperr
have been removed. If needed, a new backwards compatible function to
raise floating point errors could be restored. Reason for removal:
there are no known users and the functions would have made
with np.errstate() fixes much more difficult).
(gh-23922)
The numpy/old_defines.h which was part of the API deprecated since
NumPy 1.7 has been removed. This removes macros of the form
PyArray_CONSTANT. The
replace_old_macros.sed
script may be useful to convert them to the NPY_CONSTANT version.
(gh-24011)
The legacy_inner_loop_selector member of the ufunc struct is
removed to simplify improvements to the dispatching system. There
are no known users overriding or directly accessing this member.
(gh-24271)
NPY_INTPLTR has been removed to avoid confusion (see intp
redefinition).
(gh-24888)
The advanced indexing MapIter and related API has been removed.
The (truly) public part of it was not well tested and had only one
known user (Theano). Making it private will simplify improvements to
speed up ufunc.at, make advanced indexing more maintainable, and
was important for increasing the maximum number of dimensions of
arrays to 64. Please let us know if this API is important to you so
we can find a solution together.
(gh-25138)
The NPY_MAX_ELSIZE macro has been removed, as it only ever
reflected builtin numeric types and served no internal purpose.
(gh-25149)
PyArray_REFCNT and NPY_REFCOUNT are removed. Use Py_REFCNT
instead.
(gh-25156)
PyArrayFlags_Type and PyArray_NewFlagsObject as well as
PyArrayFlagsObject are private now. There is no known use-case;
use the Python API if needed.
PyArray_MoveInto, PyArray_CastTo, PyArray_CastAnyTo are
removed use PyArray_CopyInto and if absolutely needed
PyArray_CopyAnyInto (the latter does a flat copy).
PyArray_FillObjectArray is removed, its only true use was for
implementing np.empty. Create a new empty array or use
PyArray_FillWithScalar() (decrefs existing objects).
PyArray_CompareUCS4 and PyArray_CompareString are removed. Use
the standard C string comparison functions.
PyArray_ISPYTHON is removed as it is misleading, has no known
use-cases, and is easy to replace.
PyArray_FieldNames is removed, as it is unclear what it would be
useful for. It also has incorrect semantics in some possible
use-cases.
PyArray_TypestrConvert is removed, since it seems a misnomer and
unlikely to be used by anyone. If you know the size or are limited
to few types, just use it explicitly, otherwise go via Python
strings.
(gh-25292)
PyDataType_GetDatetimeMetaData is removed, it did not actually do
anything since at least NumPy 1.7.
(gh-25802)
PyArray_GetCastFunc is removed. Note that custom legacy user
dtypes can still provide a castfunc as their implementation, but any
access to them is now removed. The reason for this is that NumPy
never used these internally for many years. If you use simple
numeric types, please just use C casts directly. In case you require
an alternative, please let us know so we can create new API such as
PyArray_CastBuffer() which could use old or new cast functions
depending on the NumPy version.
(gh-25161)

New Features

`np.add` was extended to work with `unicode` and `bytes` dtypes.

(gh-24858)

A new `bitwise_count` function

This new function counts the number of 1-bits in a number.
numpy.bitwise_count works on all the numpy integer types
and integer-like objects.

>>> a = np.array([2**i - 1 for i in range(16)])
>>> np.bitwise_count(a)
array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15],
      dtype=uint8)

(gh-19355)

macOS Accelerate support, including the ILP64

Support for the updated Accelerate BLAS/LAPACK library, including ILP64
(64-bit integer) support, in macOS 13.3 has been added. This brings
arm64 support, and significant performance improvements of up to 10x for
commonly used linear algebra operations. When Accelerate is selected at
build time, or if no explicit BLAS library selection is done, the 13.3+
version will automatically be used if available.

(gh-24053)

Binary wheels are also available. On macOS >=14.0, users who install
NumPy from PyPI will get wheels built against Accelerate rather than
OpenBLAS.

(gh-25255)

Option to use weights for quantile and percentile functions

A weights keyword is now available for numpy.quantile, numpy.percentile,
numpy.nanquantile and numpy.nanpercentile. Only method="inverted_cdf"
supports weights.

(gh-24254)

Improved CPU optimization tracking

A new tracer mechanism is available which enables tracking of the
enabled targets for each optimized function (i.e., that uses
hardware-specific SIMD instructions) in the NumPy library. With this
enhancement, it becomes possible to precisely monitor the enabled CPU
dispatch targets for the dispatched functions.

A new function named opt_func_info has been added to the new namespace
numpy.lib.introspect, offering this tracing capability. This function allows
you to retrieve information about the enabled targets based on function names
and data type signatures.

(gh-24420)

A new Meson backend for `f2py`

f2py in compile mode (i.e. f2py -c) now accepts the
--backend meson option. This is the default option for Python >=3.12.
For older Python versions, f2py will still default to
--backend distutils.

To support this in realistic use-cases, in compile mode f2py takes a
--dep flag one or many times which maps to dependency() calls in the
meson backend, and does nothing in the distutils backend.

There are no changes for users of f2py only as a code generator, i.e.
without -c.

(gh-24532)

`bind(c)` support for `f2py`

Both functions and subroutines can be annotated with bind(c). f2py
will handle both the correct type mapping, and preserve the unique label
for other C interfaces.

Note: bind(c, name = 'routine_name_other_than_fortran_routine') is
not honored by the f2py bindings by design, since bind(c) with the
name is meant to guarantee only the same name in C and Fortran, not in
Python and Fortran.

(gh-24555)

A new `strict` option for several testing functions

The strict keyword is now available for numpy.testing.assert_allclose,
numpy.testing.assert_equal, and numpy.testing.assert_array_less. Setting
strict=True will disable the broadcasting behaviour for scalars and ensure
that input arrays have the same data type.

(gh-24680,
gh-24770,
gh-24775)

Add `np.core.umath.find` and `np.core.umath.rfind` UFuncs

Add two find and rfind UFuncs that operate on unicode or byte
strings and are used in np.char. They operate similar to str.find
and str.rfind.

(gh-24868)

`diagonal` and `trace` for `numpy.linalg`

numpy.linalg.diagonal and numpy.linalg.trace have been added, which are
array API standard-compatible variants of numpy.diagonal and numpy.trace.
They differ in the default axis selection which define 2-D sub-arrays.

(gh-24887)

New `long` and `ulong` dtypes

numpy.long and numpy.ulong have been added as NumPy integers mapping to
C's long and unsigned long. Prior to NumPy 1.24, numpy.long was an alias
to Python's int.

(gh-24922)

`svdvals` for `numpy.linalg`

numpy.linalg.svdvals has been added. It computes singular values for (a stack
of) matrices. Executing np.svdvals(x) is the same as calling np.svd(x, compute_uv=False, hermitian=False). This function is compatible with the array
API standard.

(gh-24940)

A new `isdtype` function

numpy.isdtype was added to provide a canonical way to classify NumPy's
dtypes in compliance with the array API standard.

(gh-25054)

A new `astype` function

numpy.astype was added to provide an array API standard-compatible
alternative to the numpy.ndarray.astype method.

(gh-25079)

Array API compatible functions' aliases

13 aliases for existing functions were added to improve compatibility
with the array API standard:

Trigonometry: acos, acosh, asin, asinh, atan, atanh,
atan2.
Bitwise: bitwise_left_shift, bitwise_invert,
bitwise_right_shift.
Misc: concat, permute_dims, pow.
In numpy.linalg: tensordot, matmul.

(gh-25086)

New `unique_*` functions

The numpy.unique_all, numpy.unique_counts, numpy.unique_inverse, and
numpy.unique_values functions have been added. They provide functionality of
numpy.unique with different sets of flags. They are array API
standard-compatible, and because the number of arrays they return does not
depend on the values of input arguments, they are easier to target for JIT
compilation.

(gh-25088)

Matrix transpose support for ndarrays

NumPy now offers support for calculating the matrix transpose of an
array (or stack of arrays). The matrix transpose is equivalent to
swapping the last two axes of an array. Both np.ndarray and
np.ma.MaskedArray now expose a .mT attribute, and there is a
matching new numpy.matrix_transpose function.

(gh-23762)

Array API compatible functions for `numpy.linalg`

Six new functions and two aliases were added to improve compatibility
with the Array API standard for `numpy.linalg`:

numpy.linalg.matrix_norm - Computes the matrix norm of
a matrix (or a stack of matrices).
numpy.linalg.vector_norm - Computes the vector norm of
a vector (or batch of vectors).
numpy.vecdot - Computes the (vector) dot product of
two arrays.
numpy.linalg.vecdot - An alias for
numpy.vecdot.
numpy.linalg.matrix_transpose - An alias for
numpy.matrix_transpose.
(gh-25155)
numpy.linalg.outer has been added. It computes the
outer product of two vectors. It differs from
numpy.outer by accepting one-dimensional arrays only.
This function is compatible with the array API standard.
(gh-25101)
numpy.linalg.cross has been added. It computes the
cross product of two (arrays of) 3-dimensional vectors. It differs
from numpy.cross by accepting three-dimensional
vectors only. This function is compatible with the array API
standard.
(gh-25145)

A `correction` argument for `var` and `std`

A correction argument was added to numpy.var and numpy.std, which is an
array API standard compatible alternative to ddof. As both arguments serve a
similar purpose, only one of them can be provided at the same time.

(gh-25169)

`ndarray.device` and `ndarray.to_device`

An ndarray.device attribute and ndarray.to_device method were added
to numpy.ndarray for array API standard compatibility.

Additionally, device keyword-only arguments were added to:
numpy.asarray, numpy.arange, numpy.empty, numpy.empty_like,
numpy.eye, numpy.full, numpy.full_like, numpy.linspace, numpy.ones,
numpy.ones_like, numpy.zeros, and numpy.zeros_like.

For all these new arguments, only device="cpu" is supported.

(gh-25233)

StringDType has been added to NumPy

We have added a new variable-width UTF-8 encoded string data type, implementing
a "NumPy array of Python strings", including support for a user-provided
missing data sentinel. It is intended as a drop-in replacement for arrays of
Python strings and missing data sentinels using the object dtype. See
NEP 55 and the documentation
of stringdtype for more details.

(gh-25347)

New keywords for `cholesky` and `pinv`

The upper and rtol keywords were added to
numpy.linalg.cholesky and numpy.linalg.pinv,
respectively, to improve array API standard compatibility.

For numpy.linalg.pinv, if neither rcond nor rtol is
specified, the rcond's default is used. We plan to deprecate and
remove rcond in the future.

(gh-25388)

New keywords for `sort`, `argsort` and `linalg.matrix_rank`

New keyword parameters were added to improve array API standard
compatibility:

rtol was added to numpy.linalg.matrix_rank.
stable was added to numpy.sort and
numpy.argsort.

(gh-25437)

New `numpy.strings` namespace for string ufuncs

NumPy now implements some string operations as ufuncs. The old np.char
namespace is still available, and where possible the string manipulation
functions in that namespace have been updated to use the new ufuncs,
substantially improving their performance.

Where possible, we suggest updating code to use functions in
np.strings instead of np.char. In the future we may deprecate
np.char in favor of np.strings.

(gh-25463)

`numpy.fft` support for different precisions and in-place calculations

The various FFT routines in numpy.fft now do their
calculations natively in float, double, or long double precision,
depending on the input precision, instead of always calculating in
double precision. Hence, the calculation will now be less precise for
single and more precise for long double precision. The data type of the
output array will now be adjusted accordingly.

Furthermore, all FFT routines have gained an out argument that can be
used for in-place calculations.

(gh-25536)

configtool and pkg-config support

A new numpy-config CLI script is available that can be queried for the
NumPy version and for compile flags needed to use the NumPy C API. This
will allow build systems to better support the use of NumPy as a
dependency. Also, a numpy.pc pkg-config file is now included with
Numpy. In order to find its location for use with PKG_CONFIG_PATH, use
numpy-config --pkgconfigdir.

(gh-25730)

Array API standard support in the main namespace

The main numpy namespace now supports the array API standard. See
array-api-standard-compatibility for
details.

(gh-25911)

Improvements

Strings are now supported by `any`, `all`, and the logical ufuncs.

(gh-25651)

Integer sequences as the shape argument for `memmap`

numpy.memmap can now be created with any integer sequence
as the shape argument, such as a list or numpy array of integers.
Previously, only the types of tuple and int could be used without
raising an error.

(gh-23729)

`errstate` is now faster and context safe

The numpy.errstate context manager/decorator is now faster
and safer. Previously, it was not context safe and had (rare) issues
with thread-safety.

(gh-23936)

AArch64 quicksort speed improved by using Highway's VQSort

The first introduction of the Google Highway library, using VQSort on
AArch64. Execution time is improved by up to 16x in some cases, see the
PR for benchmark results. Extensions to other platforms will be done in
the future.

(gh-24018)

Complex types - underlying C type changes

The underlying C types for all of NumPy's complex types have been
changed to use C99 complex types.
While this change does not affect the memory layout of complex
types, it changes the API to be used to directly retrieve or write
the real or complex part of the complex number, since direct field
access (as in c.real or c.imag) is no longer an option. You can
now use utilities provided in numpy/npy_math.h to do these
operations, like this:
```
npy_cdouble c;
npy_csetreal(&c, 1.0);
npy_csetimag(&c, 0.0);
printf("%d + %di\n", npy_creal(c), npy_cimag(c));
```
To ease cross-version compatibility, equivalent macros and a
compatibility layer have been added which can be used by downstream
packages to continue to support both NumPy 1.x and 2.x. See
complex-numbers for more info.
numpy/npy_common.h now includes complex.h, which means that
complex is now a reserved keyword.

(gh-24085)

`iso_c_binding` support and improved common blocks for `f2py`

Previously, users would have to define their own custom f2cmap file to
use type mappings defined by the Fortran2003 iso_c_binding intrinsic
module. These type maps are now natively supported by f2py

(gh-24555)

f2py now handles common blocks which have kind specifications from
modules. This further expands the usability of intrinsics like
iso_fortran_env and iso_c_binding.

(gh-25186)

Call `str` automatically on third argument to functions like `assert_equal`

The third argument to functions like
numpy.testing.assert_equal now has str called on it
automatically. This way it mimics the built-in assert statement, where
assert_equal(a, b, obj) works like assert a == b, obj.

(gh-24877)

Support for array-like `atol`/`rtol` in `isclose`, `allclose`

The keywords atol and rtol in numpy.isclose and
numpy.allclose now accept both scalars and arrays. An
array, if given, must broadcast to the shapes of the first two array
arguments.

(gh-24878)

Consistent failure messages in test functions

Previously, some numpy.testing assertions printed messages
that referred to the actual and desired results as x and y. Now,
these values are consistently referred to as ACTUAL and DESIRED.

(gh-24931)

n-D FFT transforms allow `s[i] == -1`

The numpy.fft.fftn, numpy.fft.ifftn,
numpy.fft.rfftn, numpy.fft.irfftn,
numpy.fft.fft2, numpy.fft.ifft2,
numpy.fft.rfft2 and numpy.fft.irfft2
functions now use the whole input array along the axis i if
s[i] == -1, in line with the array API standard.

(gh-25495)

Guard PyArrayScalar_VAL and PyUnicodeScalarObject for the limited API

PyUnicodeScalarObject holds a PyUnicodeObject, which is not
available when using Py_LIMITED_API. Add guards to hide it and
consequently also make the PyArrayScalar_VAL macro hidden.

(gh-25531)

Changes

np.gradient() now returns a tuple rather than a list making the
return value immutable.
(gh-23861)
Being fully context and thread-safe, np.errstate can only be
entered once now.
np.setbufsize is now tied to np.errstate(): leaving an
np.errstate context will also reset the bufsize.
(gh-23936)
A new public np.lib.array_utils submodule has been introduced and
it currently contains three functions: byte_bounds (moved from
np.lib.utils), normalize_axis_tuple and normalize_axis_index.
(gh-24540)
Introduce numpy.bool as the new canonical name for
NumPy's boolean dtype, and make numpy.bool\_ an alias
to it. Note that until NumPy 1.24, np.bool was an alias to
Python's builtin bool. The new name helps with array API standard
compatibility and is a more intuitive name.
(gh-25080)
The dtype.flags value was previously stored as a signed integer.
This means that the aligned dtype struct flag lead to negative flags
being set (-128 rather than 128). This flag is now stored unsigned
(positive). Code which checks flags manually may need to adapt. This
may include code compiled with Cython 0.29.x.
(gh-25816)

Representation of NumPy scalars changed

As per NEP 51, the scalar representation has been updated to include the type
information to avoid confusion with Python scalars.

Scalars are now printed as np.float64(3.0) rather than just 3.0.
This may disrupt workflows that store representations of numbers (e.g.,
to files) making it harder to read them. They should be stored as
explicit strings, for example by using str() or f"{scalar!s}". For
the time being, affected users can use
np.set_printoptions(legacy="1.25") to get the old behavior (with
possibly a few exceptions). Documentation of downstream projects may
require larger updates, if code snippets are tested. We are working on
tooling for
doctest-plus
to facilitate updates.

(gh-22449)

Truthiness of NumPy strings changed

NumPy strings previously were inconsistent about how they defined if the
string is True or False and the definition did not match the one
used by Python. Strings are now considered True when they are
non-empty and False when they are empty. This changes the following
distinct cases:

Casts from string to boolean were previously roughly equivalent to
string_array.astype(np.int64).astype(bool), meaning that only
valid integers could be cast. Now a string of "0" will be
considered True since it is not empty. If you need the old
behavior, you may use the above step (casting to integer first) or
string_array == "0" (if the input is only ever 0 or 1). To get
the new result on old NumPy versions use string_array != "".
np.nonzero(string_array) previously ignored whitespace so that a
string only containing whitespace was considered False. Whitespace
is now considered True.

This change does not affect np.loadtxt, np.fromstring, or
np.genfromtxt. The first two still use the integer definition, while
genfromtxt continues to match for "true" (ignoring case). However,
if np.bool_ is used as a converter the result will change.

The change does affect np.fromregex as it uses direct assignments.

(gh-23871)

A `mean` keyword was added to var and std function

Often when the standard deviation is needed the mean is also needed. The
same holds for the variance and the mean. Until now the mean is then
calculated twice, the change introduced here for the numpy.var and
numpy.std functions allows for passing in a precalculated mean as an keyword
argument. See the docstrings for details and an example illustrating the
speed-up.

(gh-24126)

Remove datetime64 deprecation warning when constructing with timezone

The numpy.datetime64 method now issues a UserWarning rather than a
DeprecationWarning whenever a timezone is included in the datetime string that
is provided.

(gh-24193)

Default integer dtype is now 64-bit on 64-bit Windows

The default NumPy integer is now 64-bit on all 64-bit systems as the
historic 32-bit default on Windows was a common source of issues. Most
users should not notice this. The main issues may occur with code
interfacing with libraries written in a compiled language like C. For
more information see migration_windows_int64.

(gh-24224)

Renamed `numpy.core` to `numpy._core`

Accessing numpy.core now emits a DeprecationWarning. In practice we
have found that most downstream usage of numpy.core was to access
functionality that is available in the main numpy namespace. If for
some reason you are using functionality in numpy.core that is not
available in the main numpy namespace, this means you are likely using
private NumPy internals. You can still access these internals via
numpy._core without a deprecation warning but we do not provide any
backward compatibility guarantees for NumPy internals. Please open an
issue if you think a mistake was made and something needs to be made
public.

(gh-24634)

The "relaxed strides" debug build option, which was previously enabled
through the NPY_RELAXED_STRIDES_DEBUG environment variable or the
-Drelaxed-strides-debug config-settings flag has been removed.

(gh-24717)

Redefinition of `np.intp`/`np.uintp` (almost never a change)

Due to the actual use of these types almost always matching the use of
size_t/Py_ssize_t this is now the definition in C. Previously, it
matched intptr_t and uintptr_t which would often have been subtly
incorrect. This has no effect on the vast majority of machines since the
size of these types only differ on extremely niche platforms.

However, it means that:

Pointers may not necessarily fit into an intp typed array anymore.
The p and P character codes can still be used, however.
Creating intptr_t or uintptr_t typed arrays in C remains
possible in a cross-platform way via PyArray_DescrFromType('p').
The new character codes nN were introduced.
It is now correct to use the Python C-API functions when parsing to
npy_intp typed arguments.

(gh-24888)

`numpy.fft.helper` made private

numpy.fft.helper was renamed to numpy.fft._helper to indicate that
it is a private submodule. All public functions exported by it should be
accessed from numpy.fft.

(gh-24945)

`numpy.linalg.linalg` made private

numpy.linalg.linalg was renamed to numpy.linalg._linalg to indicate
that it is a private submodule. All public functions exported by it
should be accessed from numpy.linalg.

(gh-24946)

Out-of-bound axis not the same as `axis=None`

In some cases axis=32 or for concatenate any large value was the same
as axis=None. Except for concatenate this was deprecate. Any out of
bound axis value will now error, make sure to use axis=None.

(gh-25149)

New `copy` keyword meaning for `array` and `asarray` constructors

Now numpy.array and numpy.asarray support
three values for copy parameter:

None - A copy will only be made if it is necessary.
True - Always make a copy.
False - Never make a copy. If a copy is required a ValueError is
raised.

The meaning of False changed as it now raises an exception if a copy
is needed.

(gh-25168)

The `array` special method now takes a `copy` keyword argument.

NumPy will pass copy to the __array__ special method in situations
where it would be set to a non-default value (e.g. in a call to
np.asarray(some_object, copy=False)). Currently, if an unexpected
keyword argument error is raised after this, NumPy will print a warning
and re-try without the copy keyword argument. Implementations of
objects implementing the __array__ protocol should accept a copy
keyword argument with the same meaning as when passed to
numpy.array or numpy.asarray.

(gh-25168)

Cleanup of initialization of `numpy.dtype` with strings with commas

The interpretation of strings with commas is changed slightly, in that a
trailing comma will now always create a structured dtype. E.g., where
previously np.dtype("i") and np.dtype("i,") were treated as
identical, now np.dtype("i,") will create a structured dtype, with a
single field. This is analogous to np.dtype("i,i") creating a
structured dtype with two fields, and makes the behaviour consistent
with that expected of tuples.

At the same time, the use of single number surrounded by parenthesis to
indicate a sub-array shape, like in np.dtype("(2)i,"), is deprecated.
Instead; one should use np.dtype("(2,)i") or np.dtype("2i").
Eventually, using a number in parentheses will raise an exception, like
is the case for initializations without a comma, like
np.dtype("(2)i").

(gh-25434)

Change in how complex sign is calculated

Following the array API standard, the complex sign is now calculated as
z / |z| (instead of the rather less logical case where the sign of the
real part was taken, unless the real part was zero, in which case the
sign of the imaginary part was returned). Like for real numbers, zero is
returned if z==0.

(gh-25441)

Return types of functions that returned a list of arrays

Functions that returned a list of ndarrays have been changed to return a
tuple of ndarrays instead. Returning tuples consistently whenever a
sequence of arrays is returned makes it easier for JIT compilers like
Numba, as well as for static type checkers in some cases, to support
these functions. Changed functions are: numpy.atleast_1d, numpy.atleast_2d,
numpy.atleast_3d, numpy.broadcast_arrays, numpy.meshgrid,
numpy.ogrid, numpy.histogramdd.

`np.unique` `return_inverse` shape for multi-dimensional inputs

When multi-dimensional inputs are passed to np.unique with
return_inverse=True, the unique_inverse output is now shaped such
that the input can be reconstructed directly using
np.take(unique, unique_inverse) when axis=None, and
np.take_along_axis(unique, unique_inverse, axis=axis) otherwise.

(gh-25553,
gh-25570)

`any` and `all` return booleans for object arrays

The any and all functions and methods now return booleans also for
object arrays. Previously, they did a reduction which behaved like the
Python or and and operators which evaluates to one of the arguments.
You can use np.logical_or.reduce and np.logical_and.reduce to
achieve the previous behavior.

(gh-25712)

`np.can_cast` cannot be called on Python int, float, or complex

np.can_cast cannot be called with Python int, float, or complex
instances anymore. This is because NEP 50 means that the result of
can_cast must not depend on the value passed in. Unfortunately, for
Python scalars whether a cast should be considered "same_kind" or
"safe" may depend on the context and value so that this is currently
not implemented. In some cases, this means you may have to add a
specific path for: if type(obj) in (int, float, complex): ....

(gh-26393)

Checksums

MD5

fcda027f9735771088e607161c913094  numpy-2.0.0-cp310-cp310-macosx_10_9_x86_64.whl
1c381a5af3e6b945c6937ab3c6e2de09  numpy-2.0.0-cp310-cp310-macosx_11_0_arm64.whl
6258de3c0599f8e3674e11898f2dd71c  numpy-2.0.0-cp310-cp310-macosx_14_0_arm64.whl
aa4d28b404566dc9f5c34a31c6cd7b23  numpy-2.0.0-cp310-cp310-macosx_14_0_x86_64.whl
6b83ba81bdc750ef9924e3dc6f7c93be  numpy-2.0.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
3d129fe67d99e0aad451742abb963ffa  numpy-2.0.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
24a060577965bd2a573ed87cbd207b4c  numpy-2.0.0-cp310-cp310-musllinux_1_1_x86_64.whl
b00832f558669aacf855c4f5e9cf31d1  numpy-2.0.0-cp310-cp310-musllinux_1_2_aarch64.whl
cfe7420d294c583b90cfe07b730136dc  numpy-2.0.0-cp310-cp310-win32.whl
cff9da6b9fe5ad3b05dd3526dff00ac2  numpy-2.0.0-cp310-cp310-win_amd64.whl
f390e03564df5ea37a97ac10cf0cbb00  numpy-2.0.0-cp311-cp311-macosx_10_9_x86_64.whl
a006b081decba286a321de67a1abe246  numpy-2.0.0-cp311-cp311-macosx_11_0_arm64.whl
6aea3e8589e33349b8170524af5a2e44  numpy-2.0.0-cp311-cp311-macosx_14_0_arm64.whl
eea8146c5dc2a306333bfea1f01f7a37  numpy-2.0.0-cp311-cp311-macosx_14_0_x86_64.whl
e96c2af477c970c8ff50ecb5d1cf754f  numpy-2.0.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
d065256e02a1d410d0db2577bb8fd9a4  numpy-2.0.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
98c570b79459342c219590c5af38d527  numpy-2.0.0-cp311-cp311-musllinux_1_1_x86_64.whl
dc435751cb926f53a9fc457f35146527  numpy-2.0.0-cp311-cp311-musllinux_1_2_aarch64.whl
aaa4b435d29022ceacb4e3dcbd43d11a  numpy-2.0.0-cp311-cp311-win32.whl
9ff8be4f581d86b2f181fe905491b19b  numpy-2.0.0-cp311-cp311-win_amd64.whl
1c9519c5e6a0c5a99715e51ac3b7c932  numpy-2.0.0-cp312-cp312-macosx_10_9_x86_64.whl
b0f26e8728523d716f5165953b35244f  numpy-2.0.0-cp312-cp312-macosx_11_0_arm64.whl
029703d0ff0e96c603c91f611926ef17  numpy-2.0.0-cp312-cp312-macosx_14_0_arm64.whl
2231ecbb380c70ddf462e9671d06612c  numpy-2.0.0-cp312-cp312-macosx_14_0_x86_64.whl
4153b50c1a3647ca58f1084fcaf3e4c6  numpy-2.0.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
82cba3915234f8018fd754ffc45e95b0  numpy-2.0.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
249047dd7255a5fcf5c45614ba211e10  numpy-2.0.0-cp312-cp312-musllinux_1_1_x86_64.whl
f7581ebfe0c9d4ae4f3b6ea09c19eea7  numpy-2.0.0-cp312-cp312-musllinux_1_2_aarch64.whl
8a0dbcd919d1d959f1846a00ebb05162  numpy-2.0.0-cp312-cp312-win32.whl
22aabdfd85ed34f02a7cdacff399c5d9  numpy-2.0.0-cp312-cp312-win_amd64.whl
1fce84122c393e05b69e2ec53ecd1137  numpy-2.0.0-cp39-cp39-macosx_10_9_x86_64.whl
81e4c1152274d85813bf14814ad4d359  numpy-2.0.0-cp39-cp39-macosx_11_0_arm64.whl
5eab1a2b427b590d2bc9d8ecd330fc21  numpy-2.0.0-cp39-cp39-macosx_14_0_arm64.whl
ab967929693baf2d2bfb00c53413ad2b  numpy-2.0.0-cp39-cp39-macosx_14_0_x86_64.whl
85d2971cd78800663766f46ba312d356  numpy-2.0.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
7e831fcf9cff5317429786a3bd123671  numpy-2.0.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
03a6426ca86ad53567e3ef61bc766013  numpy-2.0.0-cp39-cp39-musllinux_1_1_x86_64.whl
b30af2d2b99468538f45e6769f9fee2b  numpy-2.0.0-cp39-cp39-musllinux_1_2_aarch64.whl
cc9a8db8d131fb5a387e2c1342ab0065  numpy-2.0.0-cp39-cp39-win32.whl
9843951308fa31c5e36c4c6a0b090308  numpy-2.0.0-cp39-cp39-win_amd64.whl
5021eb5e225bff3e05a38a565daf8852  numpy-2.0.0-pp39-pypy39_pp73-macosx_10_9_x86_64.whl
99186fe49ac7931d3e92e8993c2faa92  numpy-2.0.0-pp39-pypy39_pp73-macosx_14_0_x86_64.whl
c39f0ab6e07d42708550899951b852b8  numpy-2.0.0-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
cbf151633948e90c93dd988777750961  numpy-2.0.0-pp39-pypy39_pp73-win_amd64.whl
a180aaba9982c6e15da6db62dab5eb4e  numpy-2.0.0.tar.gz

SHA256

04494f6ec467ccb5369d1808570ae55f6ed9b5809d7f035059000a37b8d7e86f  numpy-2.0.0-cp310-cp310-macosx_10_9_x86_64.whl
2635dbd200c2d6faf2ef9a0d04f0ecc6b13b3cad54f7c67c61155138835515d2  numpy-2.0.0-cp310-cp310-macosx_11_0_arm64.whl
0a43f0974d501842866cc83471bdb0116ba0dffdbaac33ec05e6afed5b615238  numpy-2.0.0-cp310-cp310-macosx_14_0_arm64.whl
8d83bb187fb647643bd56e1ae43f273c7f4dbcdf94550d7938cfc32566756514  numpy-2.0.0-cp310-cp310-macosx_14_0_x86_64.whl
79e843d186c8fb1b102bef3e2bc35ef81160ffef3194646a7fdd6a73c6b97196  numpy-2.0.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
6d7696c615765091cc5093f76fd1fa069870304beaccfd58b5dcc69e55ef49c1  numpy-2.0.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
b4c76e3d4c56f145d41b7b6751255feefae92edbc9a61e1758a98204200f30fc  numpy-2.0.0-cp310-cp310-musllinux_1_1_x86_64.whl
acd3a644e4807e73b4e1867b769fbf1ce8c5d80e7caaef0d90dcdc640dfc9787  numpy-2.0.0-cp310-cp310-musllinux_1_2_aarch64.whl
cee6cc0584f71adefe2c908856ccc98702baf95ff80092e4ca46061538a2ba98  numpy-2.0.0-cp310-cp310-win32.whl
ed08d2703b5972ec736451b818c2eb9da80d66c3e84aed1deeb0c345fefe461b  numpy-2.0.0-cp310-cp310-win_amd64.whl
ad0c86f3455fbd0de6c31a3056eb822fc939f81b1618f10ff3406971893b62a5  numpy-2.0.0-cp311-cp311-macosx_10_9_x86_64.whl
e7f387600d424f91576af20518334df3d97bc76a300a755f9a8d6e4f5cadd289  numpy-2.0.0-cp311-cp311-macosx_11_0_arm64.whl
34f003cb88b1ba38cb9a9a4a3161c1604973d7f9d5552c38bc2f04f829536609  numpy-2.0.0-cp311-cp311-macosx_14_0_arm64.whl
b6f6a8f45d0313db07d6d1d37bd0b112f887e1369758a5419c0370ba915b3871  numpy-2.0.0-cp311-cp311-macosx_14_0_x86_64.whl
5f64641b42b2429f56ee08b4f427a4d2daf916ec59686061de751a55aafa22e4  numpy-2.0.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
a7039a136017eaa92c1848152827e1424701532ca8e8967fe480fe1569dae581  numpy-2.0.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
46e161722e0f619749d1cd892167039015b2c2817296104487cd03ed4a955995  numpy-2.0.0-cp311-cp311-musllinux_1_1_x86_64.whl
0e50842b2295ba8414c8c1d9d957083d5dfe9e16828b37de883f51fc53c4016f  numpy-2.0.0-cp311-cp311-musllinux_1_2_aarch64.whl
2ce46fd0b8a0c947ae047d222f7136fc4d55538741373107574271bc00e20e8f  numpy-2.0.0-cp311-cp311-win32.whl
fbd6acc766814ea6443628f4e6751d0da6593dae29c08c0b2606164db026970c  numpy-2.0.0-cp311-cp311-win_amd64.whl
354f373279768fa5a584bac997de6a6c9bc535c482592d7a813bb0c09be6c76f  numpy-2.0.0-cp312-cp312-macosx_10_9_x86_64.whl
4d2f62e55a4cd9c58c1d9a1c9edaedcd857a73cb6fda875bf79093f9d9086f85  numpy-2.0.0-cp312-cp312-macosx_11_0_arm64.whl
1e72728e7501a450288fc8e1f9ebc73d90cfd4671ebbd631f3e7857c39bd16f2  numpy-2.0.0-cp312-cp312-macosx_14_0_arm64.whl
84554fc53daa8f6abf8e8a66e076aff6ece62de68523d9f665f32d2fc50fd66e  numpy-2.0.0-cp312-cp312-macosx_14_0_x86_64.whl
c73aafd1afca80afecb22718f8700b40ac7cab927b8abab3c3e337d70e10e5a2  numpy-2.0.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
49d9f7d256fbc804391a7f72d4a617302b1afac1112fac19b6c6cec63fe7fe8a  numpy-2.0.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
0ec84b9ba0654f3b962802edc91424331f423dcf5d5f926676e0150789cb3d95  numpy-2.0.0-cp312-cp312-musllinux_1_1_x86_64.whl
feff59f27338135776f6d4e2ec7aeeac5d5f7a08a83e80869121ef8164b74af9  numpy-2.0.0-cp312-cp312-musllinux_1_2_aarch64.whl
c5a59996dc61835133b56a32ebe4ef3740ea5bc19b3983ac60cc32be5a665d54  numpy-2.0.0-cp312-cp312-win32.whl
a356364941fb0593bb899a1076b92dfa2029f6f5b8ba88a14fd0984aaf76d0df  numpy-2.0.0-cp312-cp312-win_amd64.whl
e61155fae27570692ad1d327e81c6cf27d535a5d7ef97648a17d922224b216de  numpy-2.0.0-cp39-cp39-macosx_10_9_x86_64.whl
4554eb96f0fd263041baf16cf0881b3f5dafae7a59b1049acb9540c4d57bc8cb  numpy-2.0.0-cp39-cp39-macosx_11_0_arm64.whl
903703372d46bce88b6920a0cd86c3ad82dae2dbef157b5fc01b70ea1cfc430f  numpy-2.0.0-cp39-cp39-macosx_14_0_arm64.whl
3e8e01233d57639b2e30966c63d36fcea099d17c53bf424d77f088b0f4babd86  numpy-2.0.0-cp39-cp39-macosx_14_0_x86_64.whl
1cde1753efe513705a0c6d28f5884e22bdc30438bf0085c5c486cdaff40cd67a  numpy-2.0.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
821eedb7165ead9eebdb569986968b541f9908979c2da8a4967ecac4439bae3d  numpy-2.0.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
9a1712c015831da583b21c5bfe15e8684137097969c6d22e8316ba66b5baabe4  numpy-2.0.0-cp39-cp39-musllinux_1_1_x86_64.whl
9c27f0946a3536403efb0e1c28def1ae6730a72cd0d5878db38824855e3afc44  numpy-2.0.0-cp39-cp39-musllinux_1_2_aarch64.whl
63b92c512d9dbcc37f9d81b123dec99fdb318ba38c8059afc78086fe73820275  numpy-2.0.0-cp39-cp39-win32.whl
3f6bed7f840d44c08ebdb73b1825282b801799e325bcbdfa6bc5c370e5aecc65  numpy-2.0.0-cp39-cp39-win_amd64.whl
9416a5c2e92ace094e9f0082c5fd473502c91651fb896bc17690d6fc475128d6  numpy-2.0.0-pp39-pypy39_pp73-macosx_10_9_x86_64.whl
17067d097ed036636fa79f6a869ac26df7db1ba22039d962422506640314933a  numpy-2.0.0-pp39-pypy39_pp73-macosx_14_0_x86_64.whl
38ecb5b0582cd125f67a629072fed6f83562d9dd04d7e03256c9829bdec027ad  numpy-2.0.0-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
cef04d068f5fb0518a77857953193b6bb94809a806bd0a14983a8f12ada060c9  numpy-2.0.0-pp39-pypy39_pp73-win_amd64.whl
cf5d1c9e6837f8af9f92b6bd3e86d513cdc11f60fd62185cc49ec7d1aba34864  numpy-2.0.0.tar.gz

numpy 2.0.0 on Python PyPI

NumPy 2.0.0 Release Notes

Highlights

NumPy 2.0 Python API removals

__array_prepare__ is removed

Deprecations

numpy.fft deprecations for n-D transforms with None values in arguments

np.linalg.lstsq now defaults to a new rcond value

Expired deprecations

Compatibility notes

loadtxt and genfromtxt default encoding changed

f2py compatibility notes

Minor changes in behavior of sorting functions

Removed ambiguity when broadcasting in np.solve

Modified representation for Polynomial

C API changes

Datetime functionality exposed in the C API and Cython bindings

Const correctness for the generalized ufunc C API

Larger NPY_MAXDIMS and NPY_MAXARGS, NPY_RAVEL_AXIS introduced

NPY_MAXARGS not constant and PyArrayMultiIterObject size change

Required changes for custom legacy user dtypes

New Public DType API

New C-API import functions

Structured dtype information access through functions

Descriptor elsize and alignment access

NumPy 2.0 C API removals

New Features

np.add was extended to work with unicode and bytes dtypes.

A new bitwise_count function

macOS Accelerate support, including the ILP64

Option to use weights for quantile and percentile functions

Improved CPU optimization tracking

A new Meson backend for f2py

bind(c) support for f2py

A new strict option for several testing functions

Add np.core.umath.find and np.core.umath.rfind UFuncs

diagonal and trace for numpy.linalg

New long and ulong dtypes

svdvals for numpy.linalg

A new isdtype function

A new astype function