Hi,
I have v1 and v2 two vector:
v1=torch.tensor([1 ,0 ,0 ,0])
v2=torch.tensor([1, 0, 0, 0])
The resultat will be v3=v1xv2=[[1 0 ,0 ,0],[0,0,0,0],[0,0,0,0],[0,0,0,0]] ,v3 de size(4x4).
hawo I implement this in PyTorch.
Thanks
Hi,
I have v1 and v2 two vector:
v1=torch.tensor([1 ,0 ,0 ,0])
v2=torch.tensor([1, 0, 0, 0])
The resultat will be v3=v1xv2=[[1 0 ,0 ,0],[0,0,0,0],[0,0,0,0],[0,0,0,0]] ,v3 de size(4x4).
hawo I implement this in PyTorch.
Thanks
Hi Gues!
You are asking for the outer product of two vectors. pytorch has this
function under the name ger()
. So try:
torch.ger (v1, v2)
Best.
K. Frank
result = v1[..., None] @ v2[None, ...]
I think you should reshape your tensor first.
**convert your tensor as shape of ** (4,1
) and (1,4)
In [16]: a = torch.rand((4,1))
In [17]: b = torch.rand((4,1))
In [18]: b.T.mm(a)
Out[18]: tensor([[0.7928]])
In [19]: a.mm(b.T)
Out[19]:
tensor([[0.2572, 0.3554, 0.0888, 0.1154],
[0.2528, 0.3493, 0.0873, 0.1134],
[0.4203, 0.5808, 0.1451, 0.1886],
[0.0919, 0.1270, 0.0317, 0.0412]])