Data Parallel Extension for NumPy
This release will require DPC++ 2024.1.0
, which no longer supports Intel Gen9 integrated GPUs found in Intel CPUs of 10th generation and older.
dpnp.tensordot
function #1699
dpnp.nanmean
and dpnp.nanstd
functions #1654
dpnp.angle
function #1650
dpnp.logsumexp
and dpnp.reduce_hypot
functions #1648
dpnp.column_stack
, dpnp.dstack
and dpnp.row_stack
functions #1647
dpnp.nanargmax
, dpnp.nanargmin
, dpnp.nanmax
and dpnp.nanmin
functions #1646
dpnp.clip
function, available as well as a method of dpnp array #1645
dpnp.copysign
and dpnp.rsqrt
functions #1624
dpnp.linalg.slogdet
function #1607
dpnp.can_cast
function #1600
dpnp.linalg.solve
function #1598
dpnp.broadcast_arrays
function #1594
dpnp.tile
function #1586
dpnp.iinfo
and dpnp.finfo
functions #1582
dpnp.logaddexp
function #1561
dpnp.positive
function #1559
ValueError
to NotImplementedError
for unsupporting keyword arguments in array creation functions #1695
1.26.4
#1690
dpnp.true_divide
as an alias on dpnp.divide
function #1641
dpnp.vdot
function #1692
dpnp.linalg.qr
function #1673
dpnp.dot
function #1669
dpnp.linalg.inv
function #1665
dpnp.sort
and dpnp.argsort
functions, as well as implementing support of axis
keyword #1660
dpnp.linalg.cholesky
function, as well as implementing support of upper
keyword #1638
dpnp.diff
, as well as implementing support of prepend
and append
keywords #1637
dpnp.matmul
function #1616
dpnp.linalg.det
function #1607
dpnp.linalg.svd
function, as well as implementing support of full_matrices
, compute_uv
and hermitian
keywords #1604
dpnp.put_along_axis
and dpnp.take_along_axis
functions, as well as available values of axis
keyword #1636
keepdims
, initial
and where
keywords to dpnp.amax
and dpnp.amin
functions #1639
dpnp.meshgrid
function to support sparse
and copy
keyword arguments #1675
dpnp.average
function to support axis
, weights
, returned
and keepdims
keywords and dpnp.nansum
function with axis
, dtype
, keepdims
and out
keyword arguments #1654
dpnp.std
, dpnp.var
and nanvar
functions to support axis
, dtype
, out
and keepdims
keyword arguments #1635
dpnp.ogrid
and dpnp.mgrid
functions with support of device-aware keywords of compute follows data paradigm #1622
dpnp.indices
function to support dtype
and sparse
keyword arguments, as well as device-aware keywords of compute follows data paradigm #1622
dpnp.count_nonzero
function to support axis
and keepdims
keyword arguments #1615
dpnp.put_along_axis
and dpnp.take_along_axis
functions to support out
, dtype
and casting
keyword arguments #1608
dpnp.stack
and dpnp.concatenate
functions to support out
, dtype
and casting
keyword arguments #1608
dpnp.vstack
function to support dtype
and casting
keyword arguments #1595
dpnp.diag
, dpnp.diagflat
, dpnp.ptp
and dpnp.vander
functions with support of extra keywords to align with compute follows data paradigm #1579
dpnp.tri
and dpnp.identity
functions with support of device-aware keywords of compute follows data paradigm #1577
dpnp.divide
and dpnp.floor_divide
functions #1587
dpnp.cbrt
and dpnp.exp2
functions through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1624
dpnp.exp
, dpnp.expm1
, dpnp.log10
, dpnp.log1p
and dpnp.log2
functions through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1576
dpnp.abs
function through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1575
dpnp.hypot
function through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1560
dpctl.tensor
implementation for dpnp.reciprocal
function #1650
dpctl.tensor
implementation for dpnp.mean
function #1632
dpctl.tensor
implementation for dpnp.repeat
function #1614
dpctl.tensor
implementation for dpnp.argmax
and dpnp.argmin
functions #1610
dpctl.tensor
implementation for dpnp.geomspace
and dpnp.logspace
functions #1603
dpctl.tensor
implementation for dpnp.max
and dpnp.min
functions #1602
dpctl.tensor
implementation for dpnp.astype
function #1597
dpctl.tensor
implementation for dpnp.maximum
and dpnp.minimum
functions #1558
dpnp.take_along_axis
and dpnp.put_along_axis
functions #1702
dpnp.matmul
and dpnp.dot
function when out
keyword is passed #1694
dpnp.clip
where both min
and max
keywords have None
value with NumPy implementation #1670
out
keyword in elementwise functions #1656
-Wvla-extension
option enabled by default #1651
IntelDPCPPConfig.cmake
script with vendored IntelSYCLConfig.cmake
#1611
dpnp.atleast_2d
and dpnp.atleast_3d
functions and fixed to return a correct shape of resulting array #1560
dpnp.flip
, dpnp.fliplr
and dpnp.flipud
#1543
dpnp.rint
function through dpnp.round
call #1537
dpnp.real
and dpnp.imag
elementwise functions #1557
python=3.8
to dppy/label/dev
channel of Anaconda #1534
python=3.11
to dppy/label/dev
channel of Anaconda #1501
cython
support of 3.0.0
or above version #1495
Build from source
section in README.md
to state all the required prerequisite packages #1553
dpnp.hstack
and dpnp.atleast_1d
through existing functions to get rid of falling back on NumPy #1544
dpnp.asfarray
through existing functions to get rid of falling back on NumPy #1542
raw
to multi_ptr
with address_space_cast
to adopt towards changes introduced in SYCL 2020
#1538
pip
#1531
dpnp.copyto
through existing functions instead of a separate kernel #1516
dpctl.tensor
implementation #1545
dpnp.signbit
and dpnp.proj
functions implemented through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1535
dpnp.round
and dpnp.around
functions through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1520
dpnp.sign
and dpnp.negative
functions through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1523
dpnp.conjugate
and dpnp.conj
functions through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1519
dpnp.ceil
, dpnp.floor
and dpnp.trunc
functions through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1518
dpnp.remainder
and dpnp.mod
functions through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1515
dpnp.power
function through pybind11 extension of OneMKL calls where possible or leveraging on dpctl.tensor
implementation #1476
dpctl.tensor
implementation for dpnp.put
function #1529
dpctl.tensor
implementation for dpnp.roll
and dpnp.rollaxis
functions #1517
dpctl.tensor
implementation for dpnp.copy
function #1540
dpctl.tensor
implementation for dpnp.expand_dims
and dpnp.swapaxes
functions #1532
dpctl.tensor
implementation for bitwise operations #1508
dpctl.tensor
implementation for dpnp.all
and dpnp.any
functions #1512
dpctl.tensor
implementation for dpnp.stack
function #1509
dpctl.tensor
implementation for dpnp.concatenate
function #1507
dpctl.tensor
implementation for dpnp.isnan
, dpnp.isinf
and dpnp.isfinite
functions #1504
dpctl.tensor
implementation for dpnp.take
function #1492
pybind11=2.11.1
version #1510
flake8=6.1.0
and black=23.7.0
#1505
dpnp.sum
with axis=1
#1489
Logically dead code
issue addressed by Coverity scan #1541
Arguments in wrong order
issue addressed by Coverity scan #1513
Pointer to local outside scope
issue addressed by Coverity scan #1514
dpnp.allclose
function on a device without fp64 support #1536
Full Changelog: https://github.com/IntelPython/dpnp/compare/0.12.1...0.13.0
classifiers metadata
to a description of dpnp package #1460
pygrep-hooks
to pre-commit config #1454
flake8
to pre-commit config #1453
isort
to pre-commit config #1451
clang
format to pre-commit config #1450
black
to pre-commit config #1449
pre-commit
hooks #1448
dpctl>=0.14.5
as host and run dependencies #1481
cython
package to a version less than 3.0.0
#1480
dpnp.sum
with axis=0
as a pybind11 extension #1479
dpnp.square
function through pybind11 extension of OneMKL call where possible or Leveraging on dpctl.tensor
implementation #1473
dpnp.cos
and dpnp.sin
functions through pybind11 extension of OneMKL calls where possible or Leveraging on dpctl.tensor
implementation #1471
dpnp.sqrt
function through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1470
dpnp.log
function through pybind11 extension of OneMKL call where possible or leveraging on dpctl.tensor
implementation #1469
dpctl.tensor
implementation for logical operations #1464
dpctl.tensor
implementation for dpnp.floor_divide
function #1462
dpctl.tensor
implementation for comparison functions #1458
dpnp.dot
function to support OneMKL calls for input and output arrays with strides #1477
dpnp.linalg
functions on a device without fp64 support #1474
dtype
check of fp64 support by the resulting array in call_origin
function #1457
std::getenv()
call on Windows #1452
Full Changelog: https://github.com/IntelPython/dpnp/compare/0.12.0...0.12.1
dpnp.broadcast_to
function #1333
dpnp.extract
function #1340
dpnp.linalg.eigh
function through pybind11 extension of OneMKL call #1383
dpnp.mean
function #1431
out
parameter in dpnp.add
function #1329
out
parameter in dpnp.multiply
function #1365
out
parameter in dpnp.sqrt
function #1332
rowvar
parameter in dpnp.cov
function #1371
nbytes
property to dpnp array #1359
dpctl.tensor
implementation for dpnp.place
function #1337
dpctl.tensor
implementation for dpnp.moveaxis
function #1382
dpctl.tensor
implementation for dpnp.squeeze
function #1381
dpctl.tensor
implementation for dpnp.where
function #1380
dpctl.tensor
implementation for dpnp.transpose
function #1389
dpctl.tensor
implementation for dpnp.reshape
function #1391
dpctl.tensor
implementation for dpnp.add
, dpnp.multiply
and dpnp.subtract
functions #1430
dpctl.tensor
implementation for dpnp.sum
function #1426
dpctl.tensor
implementation for dpnp.result_type
function #1435
std::nth_element
function in dpnp.partition
with 1d array #1406
dpnp.__getitem__()
and dpnp.__setitem__()
functions #1362
sysroot>=2.28
and transitioned to conda-forge
channel #1408
dpnp.divide
implementation to call div
from OneMKL for C-contiguous data or to use dpctl.tensor
library otherwise #1418
dpnp.divide
#1434
dpnp.outer
implementation through dpnp.multiply
with broadcasted arrays #1436
dpctl>=0.14.3
as host and run dependencies #1437
dpnp.cov
through existing dpnp function instead of a separate kernel #1396
dpnp.asarray
function to accept a sequence of dpnp arrays #1355
dpnp.sum
with an empty array #1369
sycl::abs
with DPC++ 2023.2.0 #1393
Full Changelog: https://github.com/IntelPython/dpnp/compare/0.11.1...0.12.0
Compatibility with dpctl version 0.12.
DPNP 0.8.0 Compatibility with dpctl version 0.10. Compatibility with numba-dppy version 0.16. Migrating to dpctl.tensor.usm_ndarray as default data container.
Backend library API changes