pytorch中的transpose()函数

torch.transpose(Tensor,dim0,dim1)是pytorch中的ndarray矩阵进行转置的操作

例如：x = ([[0,1,2],[3,4,5],[6,7,8]])

 

 

 

 

我们先把它转为矩阵

import torch
import numpy as ny
x = ([[0,1,2],[3,4,5],[6,7,8]])
x = ny.matrix(x)
print (x)

'''
[[0 1 2]
 [3 4 5]
 [6 7 8]]
'''
#默认转置
y1 = x.transpose()
print (y1)

'''
[[0 3 6]
 [1 4 7]
 [2 5 8]]
'''
#正常顺序显示
y2 = x.transpose(0,1)
print (y2)

'''
[[0 1 2]
[3 4 5]
[6 7 8]]
'''

#转置
y3 = x.transpose(1,0)
 print (y3)

''' 
[[0 3 6]
 [1 4 7]
 [2 5 8]]
 '''

如果您对转置不要理解

那你可以这样去理解他的操作

原矩阵
    0    1
x  [0 ,  0] = 0
x  [0 ,  1] = 1
x  [0 ,  2] = 2
x  [1 ,  0] = 3
x  [1 ,  1] = 4
x  [1 ,  2] = 5
x  [2 ,  0] = 6
x  [2 ,  1] = 7
x  [2 ,  2] = 8

变换后
    1    0
x  [0 ,  0] = 0
x  [1 ,  0] = 3
x  [2 ,  0] = 6
x  [0 ,  1] = 1
x  [1 ,  1] = 4
x  [2 ,  1] = 7
x  [0 ,  2] = 2
x  [1 ,  2] = 5
x  [2 ,  2] = 8