torch掩码数据操作之masked_select()函数和masked_scatter()函数

内卷焦虑人士

已于 2024-04-25 11:15:57 修改

阅读量2.3k

点赞数 2

文章标签：深度学习人工智能 pytorch python

于 2022-09-14 18:13:08 首次发布

本文链接：https://blog.csdn.net/weixin_46398647/article/details/126857664

版权

masked_select()

选择掩码数据

import torch
x = torch.tensor([[-2,-1,0],[-1,0,1],[0,1,2]])
x
tensor([[-2, -1,  0],
        [-1,  0,  1],
        [ 0,  1,  2]])
mask = x>0
mask
tensor([[False, False, False],
        [False, False,  True],
        [False,  True,  True]])
torch.masked_select(x,mask)
tensor([1, 1, 2])

masked_scatter()

替换掩码数据

variable = torch.tensor([[100,100,100],[100,100,100],[100,100,100]])
variable                                                            
tensor([[100, 100, 100],
        [100, 100, 100],
        [100, 100, 100]])
x.masked_scatter_(mask,variable)                                    
tensor([[ -2,  -1,   0],
        [ -1,   0, 100],
        [  0, 100, 100]])

另一种方式

variable_new = torch.tensor([100,100,100])                          
variable_new                              
tensor([100, 100, 100])
x.masked_scatter_(mask,variable_new)      
tensor([[ -2,  -1,   0],
        [ -1,   0, 100],
        [  0, 100, 100]])

如果变量长度小于掩码长度，则无法替换

variable_current = torch.tensor([100,100])        
variable_current                          
tensor([100, 100])
x.masked_scatter_(mask,variable_current)  
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
RuntimeError: Number of elements of source < number of ones in mask