How to extract a row of tensor in pytorch
This article mainly introduces how to extract a tensor line from pytorch, which has a certain reference value, and interested friends can refer to it. I hope you will gain a lot after reading this article.
The experimental code is as follows:
B=torch.Tensor ([10jue 2], [4jre 5], [7je 8], [[1jue 2], [4je 5], [7je 8]) print (b.size ()) (1jue 2,3,2) print (b […) , 0]) tensor ([10, 4, 7.], [1, 4, 7.]) print (b [... , 0]. Size () (1,2,3) print (b [... , 2]) Traceback (most recent call last): File ", line 1, inIndexError: index 2 is out of bounds for dimension 3 with size 2print (b [0, …]) Tensor ([[10, 2.], [4, 5.], [7, 8.]], [[1, 2.], [4, 5.], [7, 8.]]) print (b [0,...] .size () (2,3,2) print (b [0, … , 0]. Size () (2,3) print (b [0,... , 0]) tensor ([[10., 4., 7.], [1., 4., 7.]])
[... , 0] means that the first column of digits on the fourth axis of tensor b is extracted to form tensor, [0,...] Indicates that the first column of digits on the first axis of tensor b is composed of tensor, [0,... , 0] indicates that the first column of numbers on the first and fourth axes of b are extracted to form a tensor.
Another phenomenon was found.
Print (b [... , 0:]) tensor ([10, 2.], [4, 5.], [7, 8.]], [[1, 2.], [4, 5.], [7, 8.]) print (b [... , 1:]) tensor ([2.], [5.], [8.]], [[2.], [5.], [8.]) print (b [... , 2:]) tensor ([], size= (1,2,3,0))
Supplement: use cases of [..., 0] in PyTorch
1. Example 1import torcha = torch.rand ((17,24,8)) b = a [..., 0] c = a [:,:, 0] print (b = = c)
The result is True
two。 Example 2import torcha = torch.rand ((64,17,24,8)) b = a [..., 0] c = a [:, 0] print (b = = c)
The result is True
3. Conclusion
You can see [... , 0] is equivalent to [:,:,... :, 0]
Thank you for reading this article carefully. I hope the article "how to extract a tensor Line from pytorch" shared by the editor will be helpful to everyone. At the same time, I also hope that you will support and pay attention to the industry information channel. More related knowledge is waiting for you to learn!
