mindspore.ops.ScatterNdMax
- class mindspore.ops.ScatterNdMax(use_locking=False)[source]
Applies sparse maximum to individual values or slices in a tensor.
Using given values to update parameter value through the maximum operation, along with the input indices. This operation outputs the input_x after the update is done, which makes it convenient to use the updated value.
Refer to
mindspore.ops.scatter_nd_max()
for more details.- Parameters
use_locking (bool, optional) – Whether to protect the assignment by a lock. Default:
False
.
- Inputs:
input_x (Union[Parameter, Tensor]) - The target tensor, with data type of Parameter or Tensor.
indices (Tensor) - The index to do maximum operation whose data type must be int32 or int64. The rank of indices must be at least 2 and indices.shape[-1] <= len(shape).
updates (Tensor) - The tensor to do the max operation with input_x. The data type is same as input_x, and the shape is indices.shape[:-1] + x.shape[indices.shape[-1]:].
- Outputs:
Tensor, the updated input_x, has the same shape and type as input_x.
- Supported Platforms:
Ascend
GPU
CPU
Examples
>>> import mindspore >>> import numpy as np >>> from mindspore import Tensor, ops, Parameter >>> input_x = Parameter(Tensor(np.array([1, 2, 3, 4, 5, 6, 7, 8]), mindspore.float32), name="x") >>> indices = Tensor(np.array([[2], [4], [1], [7]]), mindspore.int32) >>> updates = Tensor(np.array([6, 7, 8, 9]), mindspore.float32) >>> scatter_nd_max = ops.ScatterNdMax() >>> output = scatter_nd_max(input_x, indices, updates) >>> print(output) [ 1. 8. 6. 4. 7. 6. 7. 9.] >>> input_x = Parameter(Tensor(np.ones((4, 4, 4)), mindspore.int32)) >>> indices = Tensor(np.array([[0], [2]]), mindspore.int32) >>> updates = Tensor(np.array([[[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) >>> scatter_nd_max = ops.ScatterNdMax() >>> output = scatter_nd_max(input_x, indices, updates) >>> print(output) [[[1 1 1 1] [2 2 2 2] [3 3 3 3] [4 4 4 4]] [[1 1 1 1] [1 1 1 1] [1 1 1 1] [1 1 1 1]] [[5 5 5 5] [6 6 6 6] [7 7 7 7] [8 8 8 8]] [[1 1 1 1] [1 1 1 1] [1 1 1 1] [1 1 1 1]]]