Python 如何使用 torch.stack 功能
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/52288635/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
How to use torch.stack function
提问by ???
I have a question about torch.stack
我有一个关于 torch.stack 的问题
I have 2 tensors, a.shape=(2, 3, 4) and b.shape=(2, 3). How to stack themwithout in-place operation?
我有 2 个张量，a.shape=(2, 3, 4) 和 b.shape=(2, 3)。 如何在没有就地操作的情况下堆叠它们？
回答by arjoonn
Stacking requires same number of dimensions. One way would be to unsqueeze and stack. For example:
堆叠需要相同数量的维度。一种方法是解压和堆叠。例如：
a.size() # 2, 3, 4
b.size() # 2, 3
b = torch.unsqueeze(b, dim=2) # 2, 3, 1
# torch.unsqueeze(b, dim=-1) does the same thing
torch.stack([a, b], dim=2) # 2, 3, 5
回答by gil.fernandes
Using pytorch 1.2 or 1.4 arjoonn's answer did not work for me.
使用 pytorch 1.2 或 1.4 arjoonn 的答案对我不起作用。
Instead of torch.stackI have used torch.catwith pytorch 1.2 and 1.4:
而不是torch.stack我使用torch.catpytorch 1.2 和 1.4：
>>> import torch
>>> a = torch.randn([2, 3, 4])
>>> b = torch.randn([2, 3])
>>> b = b.unsqueeze(dim=2)
>>> b.shape
torch.Size([2, 3, 1])
>>> torch.cat([a, b], dim=2).shape
torch.Size([2, 3, 5])
If you want to use torch.stackthe dimensions of the tensors have to be the same:
如果要使用torch.stack张量的尺寸必须相同：
>>> a = torch.randn([2, 3, 4])
>>> b = torch.randn([2, 3, 4])
>>> torch.stack([a, b]).shape
torch.Size([2, 2, 3, 4])
Here is another example:
这是另一个例子：
>>> t = torch.tensor([1, 1, 2])
>>> stacked = torch.stack([t, t, t], dim=0)
>>> t.shape, stacked.shape, stacked
(torch.Size([3]),
torch.Size([3, 3]),
tensor([[1, 1, 2],
[1, 1, 2],
[1, 1, 2]]))
With stackyou have the dimparameter which lets you specify on which dimension you stack the tensors with equal dimensions.
有了stack这个dim参数，您就可以指定在哪个维度上堆叠具有相同维度的张量。
