mindspore.ops.scatter_nd_add

mindspore.ops.scatter_nd_add(input_x, indices, updates, use_locking=False)

Perform a sparse addition update on input_x based on the specified indices and update values.

$$\text{input\_x}[\text{indices}[i,...,j]]+=\text{updates}[i,...,j]$$

Note

• Support implicit type conversion and type promotion.

• The dimension of indices is at least 2, and its shape must be indices.shape[-1] <= len(indices.shape).

• The shape of updates is indices.shape[:-1] + input_x.shape[indices.shape[-1]:].

Parameters

• input_x (Union[Parameter, Tensor]) – The input parameter or tensor.

• indices (Tensor) – The specified indices.

• updates (Tensor) – The update values.

• use_locking (bool, optional) – Whether to protect the assignment by a lock. Default: False .

Returns

Tensor

Supported Platforms:

Ascend GPU CPU

Examples

>>> import mindspore
>>> import numpy as np
>>> input_x = mindspore.Parameter(mindspore.tensor([1, 2, 3, 4, 5, 6, 7, 8],
...                               mindspore.float32), name="x")
>>> indices = mindspore.tensor([[2], [4], [1], [7]], mindspore.int32)
>>> updates = mindspore.tensor([6, 7, 8, 9], mindspore.float32)
>>> output = mindspore.ops.scatter_nd_add(input_x, indices, updates, False)
>>> print(output)
[ 1. 10.  9.  4. 12.  6.  7. 17.]
>>> input_x = mindspore.Parameter(mindspore.tensor(np.zeros((4, 4, 4)), mindspore.int32))
>>> indices = mindspore.tensor([[0], [2]], mindspore.int32)
>>> updates = mindspore.tensor([[[1, 1, 1, 1], [2, 2, 2, 2], [3, 3, 3, 3], [4, 4, 4, 4]],
...                            [[5, 5, 5, 5], [6, 6, 6, 6], [7, 7, 7, 7], [8, 8, 8, 8]]], mindspore.int32)
>>> output = mindspore.ops.scatter_nd_add(input_x, indices, updates, False)
>>> print(output)
[[[1 1 1 1]
  [2 2 2 2]
  [3 3 3 3]
  [4 4 4 4]]
 [[0 0 0 0]
  [0 0 0 0]
  [0 0 0 0]
  [0 0 0 0]]
 [[5 5 5 5]
  [6 6 6 6]
  [7 7 7 7]
  [8 8 8 8]]
 [[0 0 0 0]
  [0 0 0 0]
  [0 0 0 0]
  [0 0 0 0]]]