Find the biggest of two pytorch tensor on size

#1

How to find the biggest of two pytorch tensors on size

>>> tensor1 = torch.empty(0)
>>> tensor2 = torch.empty(1)
>>> tensor1
tensor([])
>>> tensor2
tensor([5.9555e-34])

torch.maximum is returrning the empty tensor as the biggest tensor

>>> torch.maximum(tensor1,tensor2)
tensor([])

Is there a way to find the biggest tensor among two tensors (mostly 1d), base on the number of elements in the tensor.
desires output is

>>> maxfunc(tensor1,tensor2)
tensor([5.9555e-34])
#2

Something like this should work:

def maxfunc(*tensors):
    idx = torch.argmax(torch.tensor([t.nelement() for t in tensors]))
    out = tensors[idx]
    return out


tensor1 = torch.empty(0)
tensor2 = torch.empty(1)
tensor3 = torch.empty(10)

maxfunc(tensor1, tensor2)
maxfunc(tensor1, tensor2, tensor3)
maxfunc(tensor3, tensor1, tensor2)
maxfunc(tensor2)
#3

Thanks @ ptrblck, is there any performance difference if i use len(tensor) or tensor.shape[0] (1d tensor)?

#4

I don’t know if there is an expected performance difference (but you could certainly profile it).
However, these operations are also not used in my code.

#5

Thanks for the help.