I'm studying about pytorch recently. But this problem is so weird..
x=np.arrage(24)
ft=torch.FloatTensor(x)
print(floatT.view([@1])[@2])
Answer = tensor([[13., 16.], [19., 22.]])
Can there be indexing methods @1 and @2 that satisfy the Answer?
It is easier to think about if you first grab the values you care about and only then use view
to interpret it as a matrix:
# setting up
>>> import torch
>>> import numpy as np
>>> x=np.arange(24) + 3 # just to visualize the difference between indices and values
>>> x
array([ 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
20, 21, 22, 23, 24, 25, 26])
# taking the values you want and viewing as matrix
>>> ft = torch.FloatTensor(x)
>>> ft[[13, 16, 19, 22]]
tensor([16., 19., 22., 25.])
>>> ft[[13, 16, 19, 22]].view(2,2)
tensor([[16., 19.],
[22., 25.]])
by view
ing ft
as a tensor with 6 columns:
ft.view(-1, 6)
Out[]:
tensor([[ 0., 1., 2., 3., 4., 5.],
[ 6., 7., 8., 9., 10., 11.],
[12., 13., 14., 15., 16., 17.],
[18., 19., 20., 21., 22., 23.]])
you place the elements ( 13
, 19
), and ( 16
, 22
) on top of each other. Now you only need to pick them up from the right rows/columns:
.view(-1, 6)[2:, (1, 4)]
Out[]:
tensor([[13., 16.],
[19., 22.]])
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.