doc/source/release/1.17.0-notes.rst
.. currentmodule:: numpy
This NumPy release contains a number of new features that should substantially improve its performance and usefulness, see Highlights below for a summary. The Python versions supported are 3.5-3.7, note that Python 2.7 has been dropped. Python 3.8b2 should work with the released source packages, but there are no future guarantees.
Downstream developers should use Cython >= 0.29.11 for Python 3.8 support and OpenBLAS >= 3.7 (not currently out) to avoid problems on the Skylake architecture. The NumPy wheels on PyPI are built from the OpenBLAS development branch in order to avoid those problems.
A new extensible random module along with four selectable random number generators <random.BitGenerators> and improved seeding designed for use in parallel
processes has been added. The currently available bit generators are MT19937 <random.mt19937.MT19937>, PCG64 <random.pcg64.PCG64>, Philox <random.philox.Philox>, and SFC64 <random.sfc64.SFC64>. See below under
New Features.
NumPy's FFT <fft> implementation was changed from fftpack to pocketfft,
resulting in faster, more accurate transforms and better handling of datasets
of prime length. See below under Improvements.
New radix sort and timsort sorting methods. It is currently not possible to
choose which will be used. They are hardwired to the datatype and used
when either stable or mergesort is passed as the method. See below
under Improvements.
Overriding numpy functions is now possible by default,
see __array_function__ below.
numpy.errstate is now also a function decoratornumpy.polynomial functions warn when passed float in place of intPreviously functions in this module would accept float values provided they
were integral (1.0, 2.0, etc). For consistency with the rest of numpy,
doing so is now deprecated, and in future will raise a TypeError.
Similarly, passing a float like 0.5 in place of an integer will now raise a
TypeError instead of the previous ValueError.
numpy.distutils.exec_command and temp_file_nameThe internal use of these functions has been refactored and there are better
alternatives. Replace exec_command with subprocess.Popen and
temp_file_name <numpy.distutils.exec_command> with tempfile.mkstemp.
When an array is created from the C-API to wrap a pointer to data, the only
indication we have of the read-write nature of the data is the writeable
flag set during creation. It is dangerous to force the flag to writeable.
In the future it will not be possible to switch the writeable flag to True
from python.
This deprecation should not affect many users since arrays created in such
a manner are very rare in practice and only available through the NumPy C-API.
numpy.nonzero should no longer be called on 0d arraysThe behavior of numpy.nonzero on 0d arrays was surprising, making uses of it
almost always incorrect. If the old behavior was intended, it can be preserved
without a warning by using nonzero(atleast_1d(arr)) instead of
nonzero(arr). In a future release, it is most likely this will raise a
ValueError.
numpy.broadcast_arrays will warnCommonly numpy.broadcast_arrays returns a writeable array with internal
overlap, making it unsafe to write to. A future version will set the
writeable flag to False, and require users to manually set it to
True if they are sure that is what they want to do. Now writing to it will
emit a deprecation warning with instructions to set the writeable flag
True. Note that if one were to inspect the flag before setting it, one
would find it would already be True. Explicitly setting it, though, as one
will need to do in future versions, clears an internal flag that is used to
produce the deprecation warning. To help alleviate confusion, an additional
FutureWarning will be emitted when accessing the writeable flag state to
clarify the contradiction.
Note that for the C-side buffer protocol such an array will return a
readonly buffer immediately unless a writable buffer is requested. If
a writeable buffer is requested a warning will be given. When using
cython, the const qualifier should be used with such arrays to avoid
the warning (e.g. cdef const double[::1] view).
Currently, a field specified as [(name, dtype, 1)] or "1type" is
interpreted as a scalar field (i.e., the same as [(name, dtype)] or
[(name, dtype, ()]). This now raises a FutureWarning; in a future version,
it will be interpreted as a shape-(1,) field, i.e. the same as [(name, dtype, (1,))] or "(1,)type" (consistently with [(name, dtype, n)]
/ "ntype" with n>1, which is already equivalent to [(name, dtype, (n,)] / "(n,)type").
float16 subnormal roundingCasting from a different floating point precision to float16 used incorrect
rounding in some edge cases. This means in rare cases, subnormal results will
now be rounded up instead of down, changing the last bit (ULP) of the result.
Starting in version 1.12.0, numpy incorrectly returned a negatively signed zero
when using the divmod and floor_divide functions when the result was
zero. For example::
np.zeros(10)//1 array([-0., -0., -0., -0., -0., -0., -0., -0., -0., -0.])
With this release, the result is correctly returned as a positively signed zero::
np.zeros(10)//1 array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.])
MaskedArray.mask now returns a view of the mask, not the mask itselfReturning the mask itself was unsafe, as it could be reshaped in place which
would violate expectations of the masked array code. The behavior of mask <ma.MaskedArray.mask> is now consistent with data <ma.MaskedArray.data>,
which also returns a view.
The underlying mask can still be accessed with ._mask if it is needed.
Tests that contain assert x.mask is not y.mask or similar will need to be
updated.
__buffer__ attribute in numpy.frombufferLooking up __buffer__ attribute in numpy.frombuffer was undocumented and
non-functional. This code was removed. If needed, use
frombuffer(memoryview(obj), ...) instead.
out is buffered for memory overlaps in take, choose, putIf the out argument to these functions is provided and has memory overlap with the other arguments, it is now buffered to avoid order-dependent behavior.
The functions load, and lib.format.read_array take an
allow_pickle keyword which now defaults to False in response to
CVE-2019-6446 <https://nvd.nist.gov/vuln/detail/CVE-2019-6446>_.
.. currentmodule:: numpy.random
Due to bugs in the application of log to random floating point numbers,
the stream may change when sampling from ~RandomState.beta, ~RandomState.binomial,
~RandomState.laplace, ~RandomState.logistic, ~RandomState.logseries or
~RandomState.multinomial if a 0 is generated in the underlying MT19937
random stream. There is a 1 in
:math:10^{53} chance of this occurring, so the probability that the stream
changes for any given seed is extremely small. If a 0 is encountered in the
underlying generator, then the incorrect value produced (either numpy.inf or
numpy.nan) is now dropped.
.. currentmodule:: numpy
i0 now always returns a result with the same shape as the inputPreviously, the output was squeezed, such that, e.g., input with just a single
element would lead to an array scalar being returned, and inputs with shapes
such as (10, 1) would yield results that would not broadcast against the
input.
Note that we generally recommend the SciPy implementation over the numpy one: it is a proper ufunc written in C, and more than an order of magnitude faster.
can_cast no longer assumes all unsafe casting is allowedPreviously, can_cast returned True for almost all inputs for
casting='unsafe', even for cases where casting was not possible, such as
from a structured dtype to a regular one. This has been fixed, making it
more consistent with actual casting using, e.g., the .astype <ndarray.astype>
method.
ndarray.flags.writeable can be switched to true slightly more oftenIn rare cases, it was not possible to switch an array from not writeable
to writeable, although a base array is writeable. This can happen if an
intermediate ndarray.base object is writeable. Previously, only the deepest
base object was considered for this decision. However, in rare cases this
object does not have the necessary information. In that case switching to
writeable was never allowed. This has now been fixed.
npy_intp const*Previously these function arguments were declared as the more strict
npy_intp*, which prevented the caller passing constant data.
This change is backwards compatible, but now allows code like::
npy_intp const fixed_dims[] = {1, 2, 3};
// no longer complains that the const-qualifier is discarded
npy_intp size = PyArray_MultiplyList(fixed_dims, 3);
.. currentmodule:: numpy.random
numpy.random module with selectable random number generatorsA new extensible numpy.random module along with four selectable random number
generators and improved seeding designed for use in parallel processes has been
added. The currently available Bit Generators are
~mt19937.MT19937, ~pcg64.PCG64, ~philox.Philox, and ~sfc64.SFC64.
PCG64 is the new default while MT19937 is retained for backwards
compatibility. Note that the legacy random module is unchanged and is now
frozen, your current results will not change. More information is available in
the :ref:API change description <new-or-different> and in the top-level view <numpy.random> documentation.
.. currentmodule:: numpy
Support for building NumPy with the libFLAME linear algebra package as the LAPACK,
implementation, see
libFLAME <https://www.cs.utexas.edu/~flame/web/libFLAME.html>_ for details.
distutils now uses an environment variable, comma-separated and case
insensitive, to determine the detection order for BLAS libraries.
By default NPY_BLAS_ORDER=mkl,blis,openblas,atlas,accelerate,blas.
However, to force the use of OpenBLAS simply do::
NPY_BLAS_ORDER=openblas python setup.py build
which forces the use of OpenBLAS. This may be helpful for users which have a MKL installation but wishes to try out different implementations.
numpy.distutils now uses an environment variable, comma-separated and case
insensitive, to determine the detection order for LAPACK libraries.
By default NPY_LAPACK_ORDER=mkl,openblas,flame,atlas,accelerate,lapack.
However, to force the use of OpenBLAS simply do::
NPY_LAPACK_ORDER=openblas python setup.py build
which forces the use of OpenBLAS. This may be helpful for users which have a MKL installation but wishes to try out different implementations.
ufunc.reduce and related functions now accept a where maskufunc.reduce, sum, prod, min, max all
now accept a where keyword argument, which can be used to tell which
elements to include in the reduction. For reductions that do not have an
identity, it is necessary to also pass in an initial value (e.g.,
initial=np.inf for min). For instance, the equivalent of
nansum would be np.sum(a, where=~np.isnan(a)).
Both radix sort and timsort have been implemented and are now used in place of
mergesort. Due to the need to maintain backward compatibility, the sorting
kind options "stable" and "mergesort" have been made aliases of
each other with the actual sort implementation depending on the array type.
Radix sort is used for small integer types of 16 bits or less and timsort for
the remaining types. Timsort features improved performance on data containing
already or nearly sorted data and performs like mergesort on random data and
requires :math:O(n/2) working space. Details of the timsort algorithm can be
found at CPython listsort.txt <https://github.com/python/cpython/blob/3.7/Objects/listsort.txt>_.
packbits and unpackbits accept an order keywordThe order keyword defaults to big, and will order the bits
accordingly. For 'order=big' 3 will become [0, 0, 0, 0, 0, 0, 1, 1],
and [1, 1, 0, 0, 0, 0, 0, 0] for order=little
unpackbits now accepts a count parametercount allows subsetting the number of bits that will be unpacked up-front,
rather than reshaping and subsetting later, making the packbits operation
invertible, and the unpacking less wasteful. Counts larger than the number of
available bits add zero padding. Negative counts trim bits off the end instead
of counting from the beginning. None counts implement the existing behavior of
unpacking everything.
linalg.svd and linalg.pinv can be faster on hermitian inputsThese functions now accept a hermitian argument, matching the one added
to linalg.matrix_rank in 1.14.0.
timedelta64 operandsThe divmod operator now handles two timedelta64 operands, with
type signature mm->qm.
fromfile now takes an offset argumentThis function now takes an offset keyword argument for binary files,
which specifics the offset (in bytes) from the file's current position.
Defaults to 0.
padThis mode pads an array to a desired shape without initializing the new entries.
empty_like and related functions now accept a shape argumentempty_like, full_like, ones_like and zeros_like now accept a shape
keyword argument, which can be used to create a new array
as the prototype, overriding its shape as well. This is particularly useful
when combined with the __array_function__ protocol, allowing the creation
of new arbitrary-shape arrays from NumPy-like libraries when such an array
is used as the prototype.
as_integer_ratio to match the builtin floatThis returns a (numerator, denominator) pair, which can be used to construct a
fractions.Fraction.
dtype objects can be indexed with multiple fields namesarr.dtype[['a', 'b']] now returns a dtype that is equivalent to
arr[['a', 'b']].dtype, for consistency with
arr.dtype['a'] == arr['a'].dtype.
Like the dtype of structured arrays indexed with a list of fields, this dtype
has the same itemsize as the original, but only keeps a subset of the fields.
This means that arr[['a', 'b']] and arr.view(arr.dtype[['a', 'b']]) are
equivalent.
.npy files support unicode field namesA new format version of 3.0 has been introduced, which enables structured types with non-latin1 field names. This is used automatically when needed.
Error messages from array comparison tests such as
testing.assert_allclose now include "max absolute difference" and
"max relative difference," in addition to the previous "mismatch" percentage.
This information makes it easier to update absolute and relative error
tolerances.
fft module by the pocketfft libraryBoth implementations have the same ancestor (Fortran77 FFTPACK by Paul N.
Swarztrauber), but pocketfft contains additional modifications which improve
both accuracy and performance in some circumstances. For FFT lengths containing
large prime factors, pocketfft uses Bluestein's algorithm, which maintains
:math:O(N log N) run time complexity instead of deteriorating towards
:math:O(N*N) for prime lengths. Also, accuracy for real valued FFTs with near
prime lengths has improved and is on par with complex valued FFTs.
ctypes support in numpy.ctypeslibA new numpy.ctypeslib.as_ctypes_type function has been added, which can be
used to converts a dtype into a best-guess ctypes type. Thanks to this
new function, numpy.ctypeslib.as_ctypes now supports a much wider range of
array types, including structures, booleans, and integers of non-native
endianness.
numpy.errstate is now also a function decoratorCurrently, if you have a function like::
def foo():
pass
and you want to wrap the whole thing in errstate, you have to rewrite it
like so::
def foo():
with np.errstate(...):
pass
but with this change, you can do::
@np.errstate(...)
def foo():
pass
thereby saving a level of indentation
numpy.exp and numpy.log speed up for float32 implementationfloat32 implementation of exp and log now benefit from AVX2/AVX512
instruction set which are detected during runtime. exp has a max ulp
error of 2.52 and log has a max ulp error or 3.83.
numpy.padThe performance of the function has been improved for most cases by filling in a preallocated array with the desired padded shape instead of using concatenation.
numpy.interp handles infinities more robustlyIn some cases where interp would previously return nan, it now
returns an appropriate infinity.
fromfile, tofile and ndarray.dumpfromfile, ndarray.ndarray.tofile and ndarray.dump now support
the pathlib.Path type for the file/fid parameter.
isnan, isinf, and isfinite ufuncs for bool and int typesThe boolean and integer types are incapable of storing nan and inf values,
which allows us to provide specialized ufuncs that are up to 250x faster than
the previous approach.
isfinite supports datetime64 and timedelta64 typesPreviously, isfinite used to raise a TypeError on being used on these
two types.
nan_to_numnan_to_num now accepts keywords nan, posinf and neginf
allowing the user to define the value to replace the nan, positive and
negative np.inf values respectively.
Often the cause of a MemoryError is incorrect broadcasting, which results in a very large and incorrect shape. The message of the error now includes this shape to help diagnose the cause of failure.
floor, ceil, and trunc now respect builtin magic methodsThese ufuncs now call the __floor__, __ceil__, and __trunc__
methods when called on object arrays, making them compatible with
decimal.Decimal and fractions.Fraction objects.
quantile now works on fraction.Fraction and decimal.Decimal objectsIn general, this handles object arrays more gracefully, and avoids floating- point operations if exact arithmetic types are used.
matmulIt is now possible to use matmul (or the @ operator) with object arrays.
For instance, it is now possible to do::
from fractions import Fraction
a = np.array([[Fraction(1, 2), Fraction(1, 3)], [Fraction(1, 3), Fraction(1, 2)]])
b = a @ a
median and percentile family of functions no longer warn about nannumpy.median, numpy.percentile, and numpy.quantile used to emit a
RuntimeWarning when encountering an nan. Since they return the
nan value, the warning is redundant and has been removed.
timedelta64 % 0 behavior adjusted to return NaTThe modulus operation with two np.timedelta64 operands now returns
NaT in the case of division by zero, rather than returning zero
__array_function__NumPy now always checks the __array_function__ method to implement overrides
of NumPy functions on non-NumPy arrays, as described in NEP 18_. The feature
was available for testing with NumPy 1.16 if appropriate environment variables
are set, but is now always enabled.
.. _NEP 18 : http://www.numpy.org/neps/nep-0018-array-function-protocol.html
lib.recfunctions.structured_to_unstructured does not squeeze single-field viewsPreviously structured_to_unstructured(arr[['a']]) would produce a squeezed
result inconsistent with structured_to_unstructured(arr[['a', b']]). This
was accidental. The old behavior can be retained with
structured_to_unstructured(arr[['a']]).squeeze(axis=-1) or far more simply,
arr['a'].
clip now uses a ufunc under the hoodThis means that registering clip functions for custom dtypes in C via
descr->f->fastclip is deprecated - they should use the ufunc registration
mechanism instead, attaching to the np.core.umath.clip ufunc.
It also means that clip accepts where and casting arguments,
and can be override with __array_ufunc__.
A consequence of this change is that some behaviors of the old clip have
been deprecated:
nan to mean "do not clip" as one or both bounds. This didn't work
in all cases anyway, and can be better handled by passing infinities of the
appropriate sign.out argument is passed. Using
casting="unsafe" explicitly will silence this warning.Additionally, there are some corner cases with behavior changes:
max < min has changed to be more consistent across dtypes, but
should not be relied upon.min and max take part in promotion rules like they do in all
other ufuncs.__array_interface__ offset now works as documentedThe interface may use an offset value that was mistakenly ignored.
savez set to 3 for force zip64 flagsavez was not using the force_zip64 flag, which limited the size of
the archive to 2GB. But using the flag requires us to use pickle protocol 3 to
write object arrays. The protocol used was bumped to 3, meaning the archive
will be unreadable by Python2.
KeyError not ValueErrorarr['bad_field'] on a structured type raises KeyError, for consistency
with dict['bad_field'].