mindspore.ops.scatter

View Source On Gitee
mindspore.ops.scatter(input, axis, index, src)[source]

Update the value in src to input according to the specified index. Refer to mindspore.ops.tensor_scatter_elements() for more details.

Note

If src is a tensor, the backward is supported only for the case src.shape == index.shape.

Parameters

  • input (Tensor) – The input tensor.

  • axis (int) – The axis to do update operation.

  • index (Tensor) – The index to do update operation.

  • src (Tensor, float) – The data to do the update operation with input .

Returns

Tensor

Supported Platforms:

Ascend GPU CPU

Examples

>>> import mindspore
>>> input = mindspore.tensor([[1, 2, 3, 4, 5]], dtype=mindspore.float32)
>>> src = mindspore.tensor([[8, 8]], dtype=mindspore.float32)
>>> index = mindspore.tensor([[2, 4]], dtype=mindspore.int64)
>>> out = mindspore.ops.scatter(input=input, axis=1, index=index, src=src)
>>> print(out)
[[1. 2. 8. 4. 8.]]
>>> input = mindspore.tensor(mindspore.ops.zeros((5, 5)), dtype=mindspore.float32)
>>> src = mindspore.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]], dtype=mindspore.float32)
>>> index = mindspore.tensor([[0, 0, 0], [2, 2, 2], [4, 4, 4]], dtype=mindspore.int64)
>>> out = mindspore.ops.scatter(input=input, axis=0, index=index, src=src)
>>> print(out)
[[1. 2. 3. 0. 0.]
[0. 0. 0. 0. 0.]
[4. 5. 6. 0. 0.]
[0. 0. 0. 0. 0.]
[7. 8. 9. 0. 0.]]
>>> input = mindspore.tensor(mindspore.ops.zeros((5, 5)), dtype=mindspore.float32)
>>> src = mindspore.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]], dtype=mindspore.float32)
>>> index = mindspore.tensor([[0, 2, 4], [0, 2, 4], [0, 2, 4]], dtype=mindspore.int64)
>>> out = mindspore.ops.scatter(input=input, axis=1, index=index, src=src)
>>> print(out)
[[1. 0. 2. 0. 3.]
[4. 0. 5. 0. 6.]
[7. 0. 8. 0. 9.]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]]