mindspore.ops.TensorScatterSub
- class mindspore.ops.TensorScatterSub[源代码]
根据指定的更新值 input_x 和输入索引 indices,进行减法运算更新输入Tensor的值。当同一索引有不同更新值时,更新的结果将是累积减法的结果。此操作与
mindspore.ops.ScatterNdSub
类似,只是更新后的结果是通过算子output返回,而不是直接原地更新input。# 遍历所有索引 for i in range(indices.shape[0]): for j in range(indices.shape[1]): ... for k in range(indices.shape[-2]): # 最后一维是坐标维度 # 获取当前索引组合 index_tuple = (i, j, ..., k) # 获取目标位置 target_index = indices[index_tuple] # 获取对应更新值 update_value = updates[index_tuple] # 执行减法操作 output[target_index] -= update_value
更多参考详见
mindspore.ops.tensor_scatter_sub()
。- 输入:
input_x (Tensor) - 输入Tensor。 input_x 的维度必须大于等于indices.shape[-1]。
indices (Tensor) - 输入Tensor的索引,数据类型为int32或int64,rank必须大于等于2。
updates (Tensor) - 指定与 input_x 相减操作的Tensor,其数据类型与 input_x 相同。并且shape应等于
。
- 输出:
Tensor,shape和数据类型与输入 input_x 相同。
- 支持平台:
Ascend
GPU
CPU
样例:
>>> import mindspore >>> import numpy as np >>> from mindspore import Tensor, ops >>> input_x = Tensor(np.array([[-0.1, 0.3, 3.6], [0.4, 0.5, -3.2]]), mindspore.float32) >>> indices = Tensor(np.array([[0, 0], [0, 0]]), mindspore.int32) >>> updates = Tensor(np.array([1.0, 2.2]), mindspore.float32) >>> # Next, demonstrate the approximate operation process of this operator: >>> # 1, indices[0] = [0, 0], indices[1] = [0, 0] >>> # 2, And input_x[0, 0] = -0.1 >>> # 3, So input_x[indices] = [-0.1, -0.1] >>> # 4, Satisfy the above formula: input_x[indices].shape=(2) == updates.shape=(2) >>> op = ops.TensorScatterSub() >>> # 5, Perform the subtract operation for the first time: >>> # first_input_x = input_x[0][0] - updates[0] = [[-1.1, 0.3, 3.6], [0.4, 0.5, -3.2]] >>> # 6, Perform the subtract operation for the second time: >>> # second_input_x = input_x[0][0] - updates[1] = [[-3.3, 0.3, 3.6], [0.4, 0.5, -3.2]] >>> output = op(input_x, indices, updates) >>> print(output) [[-3.3000002 0.3 3.6 ] [ 0.4 0.5 -3.2 ]]