torch.triu 测试

torch.triu(input, diagonal=0, *, out=None) → Tensor
返回一个上三角矩阵
参数：
input：输入的张量
diagonal：对角线

返回矩阵(二维张量)或矩阵批次 input 的上三角部分，结果张量 out 的其他元素设置为 0。
矩阵的上三角部分定义为对角线上和之上的元素。
参数 diagonal 控制要考虑的对角线。如果 diagonal = 0，则保留主对角线之上和之上的所有元素。正值不包括主对角线上方的对角线，类似地，负值包括主对角线下方的对角线。主对角线是
的索引集 ，其中 是矩阵的维度。

diagonal=0是保留主对角线以及上面的总和
diagonal=1是保留主对角线上面一行
diagonal=1是保留主对角线上面2行

diagonal=-1是保留主对角线下面一行以及以上
diagonal=-2是保留主对角线下面2行以及以上

import torch

len_s = 7
x = torch.rand((1, len_s, len_s))
print("====>>>>>>>>>>>>>x")
print(x)
print()


a0 = torch.triu(x, diagonal=0)
a1 = torch.triu(x, diagonal=1)
a2 = torch.triu(x, diagonal=2)
a_1 = torch.triu(x, diagonal=-1)

a_2 = torch.triu(x, diagonal=-2)

print("====>>>>>>>>>>>>>a0")
print(a0)
print("====>>>>>>>>>>>>>a1")
print(a1)
print("====>>>>>>>>>>>>>a2")
print(a2)
print("====>>>>>>>>>>>>>a_1")
print(a_1)
print("====>>>>>>>>>>>>>a_2")
print(a_2)

====>>>>>>>>>>>>>x
tensor([[[0.8693, 0.2737, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.3573, 0.7882, 0.0233, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.6543, 0.3388, 0.0010, 0.7694, 0.8262, 0.3854, 0.2707],
         [0.3192, 0.3134, 0.0402, 0.6848, 0.1572, 0.2816, 0.0999],
         [0.0573, 0.7577, 0.6373, 0.5226, 0.9810, 0.1912, 0.9347],
         [0.1492, 0.9502, 0.3311, 0.2077, 0.7754, 0.9187, 0.6188],
         [0.1916, 0.2040, 0.9452, 0.0766, 0.7309, 0.5191, 0.3115]]])

====>>>>>>>>>>>>>a0
tensor([[[0.8693, 0.2737, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.0000, 0.7882, 0.0233, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.0000, 0.0000, 0.0010, 0.7694, 0.8262, 0.3854, 0.2707],
         [0.0000, 0.0000, 0.0000, 0.6848, 0.1572, 0.2816, 0.0999],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.9810, 0.1912, 0.9347],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.9187, 0.6188],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.3115]]])
====>>>>>>>>>>>>>a1
tensor([[[0.0000, 0.2737, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.0000, 0.0000, 0.0233, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.0000, 0.0000, 0.0000, 0.7694, 0.8262, 0.3854, 0.2707],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.1572, 0.2816, 0.0999],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.1912, 0.9347],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.6188],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000]]])
====>>>>>>>>>>>>>a2
tensor([[[0.0000, 0.0000, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.0000, 0.0000, 0.0000, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.8262, 0.3854, 0.2707],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.2816, 0.0999],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.9347],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.0000]]])
====>>>>>>>>>>>>>a_1
tensor([[[0.8693, 0.2737, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.3573, 0.7882, 0.0233, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.0000, 0.3388, 0.0010, 0.7694, 0.8262, 0.3854, 0.2707],
         [0.0000, 0.0000, 0.0402, 0.6848, 0.1572, 0.2816, 0.0999],
         [0.0000, 0.0000, 0.0000, 0.5226, 0.9810, 0.1912, 0.9347],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.7754, 0.9187, 0.6188],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.0000, 0.5191, 0.3115]]])
====>>>>>>>>>>>>>a_2
tensor([[[0.8693, 0.2737, 0.9769, 0.8450, 0.7284, 0.0411, 0.2800],
         [0.3573, 0.7882, 0.0233, 0.3003, 0.8225, 0.8858, 0.8372],
         [0.6543, 0.3388, 0.0010, 0.7694, 0.8262, 0.3854, 0.2707],
         [0.0000, 0.3134, 0.0402, 0.6848, 0.1572, 0.2816, 0.0999],
         [0.0000, 0.0000, 0.6373, 0.5226, 0.9810, 0.1912, 0.9347],
         [0.0000, 0.0000, 0.0000, 0.2077, 0.7754, 0.9187, 0.6188],
         [0.0000, 0.0000, 0.0000, 0.0000, 0.7309, 0.5191, 0.3115]]])

Process finished with exit code 0