array_api_negative_index package

array_api_negative_index.arange_asymmetric(stop: int, /, *, xp: ArrayNamespaceFull, dtype: Any = None, device: Any = None) → Array

Create a symmetric array with values from -stop+1 to stop-1.

Parameters:

• stop (int) – The stop value (exclusive).

• dtype (Any, optional) – The desired data type of the output array, by default None

• device (Any, optional) – The device on which to create the array, by default None

• xp (ArrayNamespaceFull, optional) – The array namespace to use, by default None

Returns:

An array [0, 1, …, stop-1, -stop+1, …, -1]. The shape is (2 * stop - 1,).

Return type:

Array

array_api_negative_index.flip_symmetric(input: Array, /, *, axis: int = 0) → Array

Flip a symmetric array.

Parameters:

• input (Array) – The input array.

• axis (int, optional) – The axis to flip, by default 0

• include_zero_twice (bool, optional) – If True, the zeroth element is included twice, by default False

Returns:

The flipped array. Forall a < input.shape[axis] result[-a-1] = result[a] = input[a] = input[-a-1]

Return type:

Array

array_api_negative_index.to_symmetric(input: Array, /, *, axis: int = 0, asymmetric: bool = False, conjugate: bool = False) → Array

Extend a tensor to its opposite symmetric form.

Parameters:

• input (Array) – The input tensor.

• axis (int, optional) – The axis to extend, by default 0

• asymmetric (bool, optional) – If True, the input tensor is multiplied by -1, by default False

• conjugate (bool, optional) – If True, the input tensor is conjugated, by default False

Returns:

The symmetric tensor. If not include_zero_twice, forall a < input.shape[axis] result[-a] = result[a] Else, forall a < input.shape[axis] result[-a-1] = result[a]

Return type:

Array