mindspore.ops.hsplit

mindspore.ops.hsplit(input, indices_or_sections)

Splits a tensor into multiple sub-tensors horizontally. It is equivalent to ops.tensor_split with \(axis=1\) .

Parameters

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

• indices_or_sections (Union[int, tuple(int), list(int)]) – See argument in mindspore.ops.tensor_split().

Returns

A list of sub-tensors.

Raises

• TypeError – If input is not Tensor.

• ValueError – If dimension of input is less than 2.

Supported Platforms:

Ascend GPU CPU

Examples

>>> input_x = np.arange(6).reshape((2, 3)).astype('float32')
>>> output = ops.hsplit(Tensor(input_x), 3)
>>> print(output)
(Tensor(shape=[2, 1], dtype=Float32, value=[[ 0.00000000e+00], [ 3.00000000e+00]]),
 Tensor(shape=[2, 1], dtype=Float32, value=[[ 1.00000000e+00], [ 4.00000000e+00]]),
 Tensor(shape=[2, 1], dtype=Float32, value=[[ 2.00000000e+00], [ 5.00000000e+00]]))