[FIXED] How can I resize a PyTorch tensor with a sliding window?

Issue

I have a tensor with size: torch.Size([118160, 1]). What I want to do is split it up into n tensors with 100 elements each, sliding by 50 elements at a time. What's the best way to achieve this with PyTorch?

Solution

A possible solution is:

window_size = 100
stride = 50
splits = [x[i:min(x.size(0),i+window_size)] for i in range(0,x.size(0),stride)]

However, the last few elements will be shorter than window_size. If this is undesired, you can do:

splits = [x[i:i+window_size] for i in range(0,x.size(0)-window_size+1,stride)]

EDIT:

A more readable solution:

# if keep_short_tails is set to True, the slices shorter than window_size at the end of the result will be kept 
def window_split(x, window_size=100, stride=50, keep_short_tails=True):
  length = x.size(0)
  splits = []

  if keep_short_tails:
    for slice_start in range(0, length, stride):
      slice_end = min(length, slice_start + window_size)
      splits.append(x[slice_start:slice_end])
  else:
    for slice_start in range(0, length - window_size + 1, stride):
      slice_end = slice_start + window_size
      splits.append(x[slice_start:slice_end])

  return splits

Answered By - Yahia Zakaria