mindspore.ops.unsorted_segment_min

mindspore.ops.unsorted_segment_min(x, segment_ids, num_segments)[source]

Computes the minimum of a tensor along segments.

The following figure shows the calculation process of unsorted_segment_min:

../../_images/UnsortedSegmentMin.png
\[\text { output }_i=\text{min}_{j \ldots} \text { data }[j \ldots]\]

where \(min\) over tuples \(j...\) such that \(segment\_ids[j...] == i\).

Note

  • If the segment_id i is absent in the segment_ids, then output[i] will be filled with the maximum value of the x's type.

  • The segment_ids must be non-negative tensor.

Parameters

  • x (Tensor) – The shape is \((x_1, x_2, ..., x_R)\). With float16, float32 or int32 data type.

  • segment_ids (Tensor) – TThe label indicates the segment to which each element belongs. Set the shape as \((x_1, x_2, ..., x_N)\), where 0 < N <= R.

  • num_segments (Union[int, Tensor], optional) – Set \(z\) as num_segments, it can be an int or 0-D Tensor.

Returns

Tensor, the shape is \((z, x_{N+1}, ..., x_R)\).

Raises

TypeError – If num_segments is not an int.

Supported Platforms:

Ascend GPU CPU

Examples

>>> from mindspore import Tensor
>>> from mindspore import ops
>>> import numpy as np
>>> x = Tensor(np.array([[1, 2, 3], [4, 5, 6], [4, 2, 1]]).astype(np.float32))
>>> segment_ids = Tensor(np.array([0, 1, 1]).astype(np.int32))
>>> num_segments = 2
>>> output = ops.unsorted_segment_min(x, segment_ids, num_segments)
>>> print(output)
[[1. 2. 3.]
 [4. 2. 1.]]