numojo.routines.manipulation

Manipulation routines (numojo.routines.manipulation)

This module implements routines that manipulate the shape and layout of arrays, such as reshaping, transposing, broadcasting, and flipping.

Functions

copyto

copyto()

ndim

Overload 1

ndim[dtype: DType](array: NDArray[dtype]) -> Int

Returns the number of dimensions of the NDArray.

Parameters:

• dtype (DType)

Args:

• array (NDArray): A NDArray.

Returns:

• Int

Overload 2

ndim[cdtype: ComplexDType](array: ComplexNDArray[cdtype]) -> Int

Returns the number of dimensions of the NDArray.

Parameters:

• cdtype (ComplexDType)

Args:

• array (ComplexNDArray): A NDArray.

Returns:

• Int

shape

Overload 1

shape[dtype: DType](array: NDArray[dtype]) -> NDArrayShape

Returns the shape of the NDArray.

Parameters:

• dtype (DType)

Args:

• array (NDArray): A NDArray.

Returns:

• NDArrayShape

Overload 2

shape[cdtype: ComplexDType](array: ComplexNDArray[cdtype]) -> NDArrayShape

Returns the shape of the NDArray.

Returns: The shape of the NDArray.

Parameters:

• cdtype (ComplexDType)

Args:

• array (ComplexNDArray): A NDArray.

Returns:

• NDArrayShape

size

Overload 1

size[dtype: DType](array: NDArray[dtype], axis: Int) -> Int

Returns the size of the NDArray.

Parameters:

• dtype (DType)

Args:

• array (NDArray): A NDArray.
• axis (Int): The axis to get the size of.

Returns:

• Int

Raises

Overload 2

size[cdtype: ComplexDType](array: ComplexNDArray[cdtype], axis: Int) -> Int

Returns the size of the NDArray.

Parameters:

• cdtype (ComplexDType)

Args:

• array (ComplexNDArray): A NDArray.
• axis (Int): The axis to get the size of.

Returns:

• Int

Raises

reshape

reshape[dtype: DType](A: NDArray[dtype], shape: NDArrayShape, order: String = "C") -> NDArray[dtype]

Returns an array of the same data with a new shape.

Parameters:

• dtype (DType)

Args:

• A (NDArray): A NDArray.
• shape (NDArrayShape): New shape.
• order (String): "C" or "F". Read in this order from the original array and write in this order into the new array.

Returns:

• NDArray

Raises

Error: If the number of elements do not match.

ravel

ravel[dtype: DType](a: NDArray[dtype], order: String = "C") -> NDArray[dtype]

Returns the raveled version of the NDArray.

Return: A contiguous flattened array.

Parameters:

• dtype (DType)

Args:

• a (NDArray): NDArray.
• order (String): The order to flatten the array.

Returns:

• NDArray

Raises

transpose

Overload 1

transpose[dtype: DType](A: NDArray[dtype], axes: List[Int]) -> NDArray[dtype]

Transpose array of any number of dimensions according to arbitrary permutation of the axes.

If axes is not given, it is equal to flipping the axes.

import numojo as nm
var A = nm.random.rand(2,3,4,5)
print(nm.transpose(A))  # A is a 4darray.
print(nm.transpose(A, axes=[3,2,1,0]))

Examples.

import numojo as nm
var arr2d = nm.random.rand(2,3)
print(nm.transpose(arr2d, axes=[0, 1]))  # equal to transpose of matrix
var arr3d = nm.random.rand(2,3,4)
print(nm.transpose(arr3d, axes=[2, 1, 0]))  # transpose 0-th and 2-th dimensions

Parameters:

• dtype (DType)

Args:

• A (NDArray)
• axes (List)

Returns:

• NDArray

Raises

Overload 2

transpose[dtype: DType](A: NDArray[dtype]) -> NDArray[dtype]

(overload) Transpose the array when axes is not given. If axes is not given, it is equal to flipping the axes. See docstring of transpose.

Parameters:

• dtype (DType)

Args:

• A (NDArray)

Returns:

• NDArray

Raises

Overload 3

transpose[dtype: DType](A: Matrix[dtype]) -> Matrix[dtype]

Transpose of matrix.

Parameters:

• dtype (DType)

Args:

• A (Matrix)

Returns:

• Matrix

reorder_layout

reorder_layout[dtype: DType](A: Matrix[dtype]) -> Matrix[dtype]

Create a new Matrix with the opposite layout from A: if A is C-contiguous, then create a new F-contiguous matrix of the same shape. If A is F-contiguous, create a new C-contiguous matrix.

Copy data into the new layout.

Parameters:

• dtype (DType)

Args:

• A (Matrix)

Returns:

• Matrix

Raises

broadcast_to

Overload 1

broadcast_to[dtype: DType](a: NDArray[dtype], shape: NDArrayShape) -> NDArray[dtype]

Parameters:

• dtype (DType)

Args:

• a (NDArray)
• shape (NDArrayShape)

Returns:

• NDArray

Raises

Overload 2

broadcast_to[dtype: DType](A: Matrix[dtype], shape: Tuple[Int, Int], override_order: String = "") -> Matrix[dtype]

Broadcasts the vector to the given shape.

Example:

> from numojo import Matrix
> a = Matrix.fromstring("1 2 3", shape=(1, 3))
> print(mat.broadcast_to(a, (3, 3)))
[[1.0   2.0     3.0]
 [1.0   2.0     3.0]
 [1.0   2.0     3.0]]
> a = Matrix.fromstring("1 2 3", shape=(3, 1))
> print(mat.broadcast_to(a, (3, 3)))
[[1.0   1.0     1.0]
 [2.0   2.0     2.0]
 [3.0   3.0     3.0]]
> a = Matrix.fromstring("1", shape=(1, 1))
> print(mat.broadcast_to(a, (3, 3)))
[[1.0   1.0     1.0]
 [1.0   1.0     1.0]
 [1.0   1.0     1.0]]
> a = Matrix.fromstring("1 2", shape=(1, 2))
> print(mat.broadcast_to(a, (1, 2)))
[[1.0   2.0]]
> a = Matrix.fromstring("1 2 3 4", shape=(2, 2))
> print(mat.broadcast_to(a, (4, 2)))
Unhandled exception caught during execution: Cannot broadcast shape 2x2 to shape 4x2!

Parameters:

• dtype (DType)

Args:

• A (Matrix)
• shape (Tuple)
• override_order (String)

Returns:

• Matrix

Raises

Overload 3

broadcast_to[dtype: DType](A: Scalar[dtype], shape: Tuple[Int, Int], order: String) -> Matrix[dtype]

Broadcasts the scalar to the given shape.

Parameters:

• dtype (DType)

Args:

• A (Scalar)
• shape (Tuple)
• order (String)

Returns:

• Matrix

Raises

flip

Overload 1

flip[dtype: DType](array: NDArray[dtype]) -> NDArray[dtype]

Returns flipped array and keep the shape.

Parameters:

• dtype (DType): DType.

Args:

• array (NDArray): A NDArray.

Returns:

• NDArray

Raises

Overload 2

flip[dtype: DType](array: NDArray[dtype], var axis: Int) -> NDArray[dtype]

Returns flipped array along the given axis.

Parameters:

• dtype (DType): DType.

Args:

• array (NDArray): A NDArray.
• axis (Int) [var]: Axis along which to flip.

Returns:

• NDArray

Raises