dpnp.dparray.dparray

class dpnp.dparray.dparray

Multi-dimensional array using USM interface for an Intel GPU device.

This class implements a subset of methods of numpy.ndarray. The difference is that this class allocates the array content useing USM interface on the current GPU device.

Parameters

shape (tuple of ints) – Length of axes.
dtype – Data type. It must be an argument of numpy.dtype.
memptr (char *) – Pointer to the array content head.
strides (tuple of ints or None) – Strides of data in memory.
order ({'C', 'F'}) – Row-major (C-style) or column-major (Fortran-style) order.

Variables

~dparray.base (None or dpnp.dparray) – Base array from which this array is created as a view.
~dparray.data (char *) – Pointer to the array content head.
~dparray.dtype (numpy.dtype) –
Dtype object of elements type.

See also

Data type objects (dtype)
~dparray.size (int) –
Number of elements this array holds.

This is equivalent to product over the shape tuple.

See also

numpy.ndarray.size

Methods

__getitem__(): Get the array item(s) x.__getitem__(key) <==> x[key]

__setitem__(): Set the array item(s) x.__setitem__(key, value) <==> x[key] = value

__len__(): Returns the size of the first dimension. Equivalent to shape[0] and also equal to size only for one-dimensional arrays.

See also

numpy.ndarray.__len__

__next__()

__iter__(): Implement iter(self).

all()

Returns True if all elements evaluate to True.

Refer to numpy.all for full documentation.

See also

numpy.all: equivalent function

any()

Returns True if any of the elements of a evaluate to True.

Refer to numpy.any for full documentation.

See also

numpy.any: equivalent function

argmax()

Returns array of indices of the maximum values along the given axis.

Parameters

axis ({None, integer}) – If None, the index is into the flattened array, otherwise along the specified axis
out ({None, array}, optional) – Array into which the result can be placed. Its type is preserved and it must be of the right shape to hold the output.

Returns

index_array

Return type

{integer_array}

Examples

>>> a = np.arange(6).reshape(2,3)
>>> a.argmax()
5
>>> a.argmax(0)
array([1, 1, 1])
>>> a.argmax(1)
array([2, 2])

argmin()

Return array of indices to the minimum values along the given axis.

Parameters

axis ({None, integer}) – If None, the index is into the flattened array, otherwise along the specified axis
out ({None, array}, optional) – Array into which the result can be placed. Its type is preserved and it must be of the right shape to hold the output.

Returns

If multi-dimension input, returns a new ndarray of indices to the minimum values along the given axis. Otherwise, returns a scalar of index to the minimum values along the given axis.

Return type

ndarray or scalar

argsort()

Return an ndarray of indices that sort the array along the specified axis.

Parameters

axis (int, optional) –
Axis along which to sort. If None, the default, the flattened array is used. .. versionchanged:: 1.13.0

Previously, the default was documented to be -1, but that was in error. At some future date, the default will change to -1, as originally intended. Until then, the axis should be given explicitly when arr.ndim > 1, to avoid a FutureWarning.
kind ({'quicksort', 'mergesort', 'heapsort', 'stable'}, optional) – The sorting algorithm used.
order (list, optional) – When a is an array with fields defined, this argument specifies which fields to compare first, second, etc. Not all fields need be specified.

Returns

index_array – Array of indices that sort a along the specified axis. In other words, a[index_array] yields a sorted a.

Return type

ndarray, int

See also

MaskedArray.sort: Describes sorting algorithms used.
dpnp.lexsort: Indirect stable sort with multiple keys.
numpy.ndarray.sort: Inplace sort.

Notes

See sort for notes on the different sorting algorithms.

astype()

Copy the array with data type casting.

Parameters

dtype – Target type.
order ({'C', 'F', 'A', 'K'}) – Row-major (C-style) or column-major (Fortran-style) order. When order is ‘A’, it uses ‘F’ if a is column-major and uses ‘C’ otherwise. And when order is ‘K’, it keeps strides as closely as possible.
copy (bool) – If it is False and no cast happens, then this method returns the array itself. Otherwise, a copy is returned.

Returns

If copy is False and no cast is required, then the array itself is returned. Otherwise, it returns a (possibly casted) copy of the array.

Note

This method currently does not support order`, casting`, copy, and subok arguments.

See also

numpy.ndarray.astype()

choose(): Construct an array from an index array and a set of arrays to choose from.

conj()

Complex-conjugate all elements.

For full documentation refer to numpy.ndarray.conj.

conjugate()

Return the complex conjugate, element-wise.

For full documentation refer to numpy.ndarray.conjugate.

copy(): Return a copy of the array.

cumprod()

cumsum()

diagonal(): Return specified diagonals.

fill()

Fill the array with a scalar value.

Parameters: value (scalar) – All elements of a will be assigned this value.

Examples

>>> a = np.array([1, 2])
>>> a.fill(0)
>>> a
array([0, 0])
>>> a = np.empty(2)
>>> a.fill(1)
>>> a
array([1.,  1.])

flatten()

Return a copy of the array collapsed into one dimension.

Parameters: order ({'C', 'F', 'A', 'K'}, optional) – ‘C’ means to flatten in row-major (C-style) order. ‘F’ means to flatten in column-major (Fortran- style) order. ‘A’ means to flatten in column-major order if a is Fortran contiguous in memory, row-major order otherwise. ‘K’ means to flatten a in the order the elements occur in memory. The default is ‘C’.
Returns: out – A copy of the input array, flattened to one dimension.
Return type: ndarray

See also

dpnp.ravel, dpnp.flat

item()

Copy an element of an array to a standard Python scalar and return it.

For full documentation refer to numpy.ndarray.item.

Examples

>>> np.random.seed(123)
>>> x = np.random.randint(9, size=(3, 3))
>>> x
array([[2, 2, 6],
       [1, 3, 6],
       [1, 0, 1]])
>>> x.item(3)
1
>>> x.item(7)
0
>>> x.item((0, 1))
2
>>> x.item((2, 2))
1

max(): Return the maximum along an axis.

mean(): Returns the average of the array elements.

min(): Return the minimum along a given axis.

partition()

Return a partitioned copy of an array. For full documentation refer to numpy.partition.

Limitations

Input array is supported as dpnp.ndarray. Input kth is supported as int. Parameters axis, kind and order are supported only with default values.

prod(): Returns the prod along a given axis.

See also

dpnp.prod for full documentation, dpnp.dparray.sum()

ptp()

ravel()

Return a contiguous flattened array.

Parameters: order ({'C', 'F', 'A', 'K'}, optional) – ‘C’ means to flatten in row-major (C-style) order. ‘F’ means to flatten in column-major (Fortran- style) order. ‘A’ means to flatten in column-major order if a is Fortran contiguous in memory, row-major order otherwise. ‘K’ means to flatten a in the order the elements occur in memory. The default is ‘C’.
Returns: out – A copy of the input array, flattened to one dimension.
Return type: ndarray

Notes

Unlike the free function dpnp.reshape, this method on dpnp.ndarray allows the elements of the shape parameter to be passed in as separate arguments. For example, a.reshape(10, 11) is equivalent to a.reshape((10, 11)).

See also

dpnp.ravel, dpnp.flat

repeat(): Repeat elements of an array.

See also

dpnp.repeat for full documentation, numpy.ndarray.repeat()

reshape(): Change the shape of the array.

See also

numpy.ndarray.reshape()

round(): Return array with each element rounded to the given number of decimals.

See also

dpnp.around for full documentation.

sort()

Sort the array

Parameters

a (array_like) – Array to be sorted.
axis (int, optional) – Axis along which to sort. If None, the array is flattened before sorting. The default is -1, which sorts along the last axis.
kind ({'quicksort', 'mergesort', 'heapsort', 'stable'}, optional) – The sorting algorithm used.
order (list, optional) – When a is a structured array, this argument specifies which fields to compare first, second, and so on. This list does not need to include all of the fields.

Returns

sorted_array – Array of the same type and shape as a.

Return type

ndarray

See also

numpy.ndarray.sort: Method to sort an array in-place.
dpnp.argsort: Indirect sort.
dpnp.lexsort: Indirect stable sort on multiple keys.
dpnp.searchsorted: Find elements in a sorted array.

Notes

See sort for notes on the different sorting algorithms.

squeeze(): Remove single-dimensional entries from the shape of an array.

See also

dpnp.squeeze for full documentation

std(): Returns the variance of the array elements, along given axis.

See also

dpnp.var for full documentation,

sum(): Returns the sum along a given axis.

See also

dpnp.sum for full documentation, dpnp.dparray.sum()

take(): Take elements from an array. For full documentation refer to numpy.take.

tobytes()

Construct Python bytes containing the raw data bytes in the array.

For full documentation refer to numpy.ndarray.tobytes.

Examples

>>> import dpnp as np
>>> x = np.array([[0, 1], [2, 3]], dtype='<u2')
>>> x.tobytes()
b'

tofile()

Write array to a file as text or binary (default).

For full documentation refer to numpy.ndarray.tofile.

tolist()

Return the array as an a.ndim-levels deep nested list of Python scalars.

For full documentation refer to numpy.ndarray.tolist.

Examples

>>> import dpnp as np
For a 1D array, ``a.tolist()`` is almost the same as ``list(a)``,
except that ``tolist`` changes numpy scalars to Python scalars:

>>> a = np.uint32([1, 2])
>>> a_list = list(a)
>>> a_list
[1, 2]
>>> type(a_list[0])
<class 'numpy.uint32'>
>>> a_tolist = a.tolist()
>>> a_tolist
[1, 2]
>>> type(a_tolist[0])
<class 'int'>

Additionally, for a 2D array, tolist applies recursively:

>>> a = np.array([[1, 2], [3, 4]])
>>> list(a)
[array([1, 2]), array([3, 4])]
>>> a.tolist()
[[1, 2], [3, 4]]

The base case for this recursion is a 0D array:

>>> a = np.array(1)
>>> list(a)
Traceback (most recent call last):
  ...
TypeError: iteration over a 0-d array
>>> a.tolist()
1

tostring()

Construct Python bytes containing the raw data bytes in the array.

This function is a compatibility alias for tobytes. Despite its name it returns bytes not strings.

For full documentation refer to numpy.ndarray.tostring.

transpose(): Returns a view of the array with axes permuted.

See also

dpnp.transpose for full documentation, numpy.ndarray.reshape()

var()

Returns the variance of the array elements along given axis.

Masked entries are ignored, and result elements which are not finite will be masked.

Refer to numpy.var for full documentation.

See also

numpy.ndarray.var: corresponding function for ndarrays
numpy.var: Equivalent function

__eq__(value, /): Return self==value.

__ne__(value, /): Return self!=value.

__lt__(value, /): Return self<value.

__le__(value, /): Return self<=value.

__gt__(value, /): Return self>value.

__ge__(value, /): Return self>=value.

Attributes

T

Shape-reversed view of the array.

If ndim < 2, then this is just a reference to the array itself.

dtype: Type of the elements in the array

flags

Object containing memory-layout information.

It only contains c_contiguous, f_contiguous, and owndata attributes. All of these are read-only. Accessing by indexes is also supported.

See also

numpy.ndarray.flags

flat: Return a flat iterator, or set a flattened version of self to value.

itemsize: Size of each element in bytes.

See also

numpy.ndarray.itemsize

nbytes

Total size of all elements in bytes.

It does not count strides or alignment of elements.

See also

numpy.ndarray.nbytes

ndim

Number of dimensions.

a.ndim is equivalent to len(a.shape).

See also

numpy.ndarray.ndim

shape

Lengths of axes. A tuple of numbers represents size of each dimention.

Setter of this property involves reshaping without copy. If the array cannot be reshaped without copy, it raises an exception.

size: Number of elements in the array.

See also

numpy.ndarray.size

strides: Strides of axes in bytes.

See also

numpy.ndarray.strides