方法原型
1torch.sort(input, dim=None, descending=False, out=None) -> (Tensor, LongTensor)
返回值
1A tuple of (sorted_tensor, sorted_indices) is returned,
2where the sorted_indices are the indices of the elements in the original input tensor.
参数
- input (Tensor) – the input tensor
形式上与 numpy.narray 类似 - dim (int, optional) – the dimension to sort along
维度，对于二维数据：dim=0 按列排序，dim=1 按行排序，默认 dim=1 - descending (bool, optional) – controls the sorting order (ascending or descending)
降序，descending=True 从大到小排序，descending=False 从小到大排序，默认 descending=Flase
实例
1import torch
2x = torch.randn(3,4)
3x #初始值，始终不变
4tensor([[-0.9950, -0.6175, -0.1253, 1.3536],
5 [ 0.1208, -0.4237, -1.1313, 0.9022],
6 [-1.1995, -0.0699, -0.4396, 0.8043]])
7sorted, indices = torch.sort(x) #按行从小到大排序
8sorted
9tensor([[-0.9950, -0.6175, -0.1253, 1.3536],
10 [-1.1313, -0.4237, 0.1208, 0.9022],
11 [-1.1995, -0.4396, -0.0699, 0.8043]])
12indices
13tensor([[0, 1, 2, 3],
14 [2, 1, 0, 3],
15 [0, 2, 1, 3]])
16sorted, indices = torch.sort(x, descending=True) #按行从大到小排序 (即反序)
17sorted
18tensor([[ 1.3536, -0.1253, -0.6175, -0.9950],
19 [ 0.9022, 0.1208, -0.4237, -1.1313],
20 [ 0.8043, -0.0699, -0.4396, -1.1995]])
21indices
22tensor([[3, 2, 1, 0],
23 [3, 0, 1, 2],
24 [3, 1, 2, 0]])
25sorted, indices = torch.sort(x, dim=0) #按列从小到大排序
26sorted
27tensor([[-1.1995, -0.6175, -1.1313, 0.8043],
28 [-0.9950, -0.4237, -0.4396, 0.9022],
29 [ 0.1208, -0.0699, -0.1253, 1.3536]])
30indices
31tensor([[2, 0, 1, 2],
32 [0, 1, 2, 1],
33 [1, 2, 0, 0]])
34sorted, indices = torch.sort(x, dim=0, descending=True) #按列从大到小排序
35sorted
36tensor([[ 0.1208, -0.0699, -0.1253, 1.3536],
37 [-0.9950, -0.4237, -0.4396, 0.9022],
38 [-1.1995, -0.6175, -1.1313, 0.8043]])
39indices
40tensor([[1, 2, 0, 0],
41 [0, 1, 2, 1],
42 [2, 0, 1, 2]])