mindspore.ops.chunk

mindspore.ops.chunk(input, chunks, axis=0)[source]

Cut the input Tensor into chunks sub-tensors along the specified axis.

Note

This function may return less then the specified number of chunks!

Parameters

  • input (Tensor) – A Tensor to be cut.

  • chunks (int) – Number of sub-tensors to cut.

  • axis (int, optional) – Specify the dimensions that you want to split. Default: 0 .

Returns

A tuple of sub-tensors.

Raises

  • TypeError – If argument input is not Tensor.

  • TypeError – The sum of chunks is not int.

  • TypeError – If argument axis is not int.

  • ValueError – If argument axis is out of range of \([-input.ndim, input.ndim)\) .

  • ValueError – If argument chunks is not positive number.

Supported Platforms:

Ascend GPU CPU

Examples

>>> import numpy as np
>>> from mindspore import ops, Tensor
>>> input_x = np.arange(9).astype("float32")
>>> output = ops.chunk(Tensor(input_x), 3)
>>> print(output)
(Tensor(shape=[3], dtype=Float32, value= [ 0.00000000e+00,  1.00000000e+00,  2.00000000e+00]),
 Tensor(shape=[3], dtype=Float32, value= [ 3.00000000e+00,  4.00000000e+00,  5.00000000e+00]),
 Tensor(shape=[3], dtype=Float32, value= [ 6.00000000e+00,  7.00000000e+00,  8.00000000e+00]))