view在pytorch中是用来改变张量的shape的,简单又好用。
pytorch中view的用法通常是直接在张量名后用.view调用,然后放入自己想要的shape。如
tensor_name.view(shape)
Example:
1. 直接用法:
>>> x = torch.randn(4, 4)
>>> x.size()
torch.Size([4, 4])
>>> y = x.view(16)
>>> y.size()
torch.Size([16])
2. 强调某一维度的尺寸:
>>> z = x.view(-1, 8) # the size -1 is inferred from other dimensions
>>> z.size()
torch.Size([2, 8])
3. 拉直张量:(直接填-1表示拉直, 等价于tensor_name.flatten())
>>> y = x.view(-1)
>>> y.size()
torch.Size([16])
4. 做维度变换时不改变内存排列
>>> a = torch.randn(1, 2, 3, 4)
>>> a.size()
torch.Size([1, 2, 3, 4])
>>> b = a.transpose(1, 2) # Swaps 2nd and 3rd dimension
>>> b.size()
torch.Size([1, 3, 2, 4])
>>> c = a.view(1, 3, 2, 4) # Does not change tensor layout in memory
>>> c.size()
torch.Size([1, 3, 2, 4])
>>> torch.equal(b, c)
False
注意最后的False,在张量b和c是不等价的。从这里我们可以看得出来,view函数如其名,只改变“看起来”的样子,不会改变张量在内存中的排列。