Most examples of pdist (or cdist) seems to be between two tensors. If I have a piecewise curve defined as follows: line=torch.tensor([(-104.6400,0.0000),(-104.6400,0.1500),(-103.5500,0.5140),(-98.1000,1.0775),(-92.6500,1.4553)]) what I have coded to get the segment lenghts is i=0 for point in line: if i>0: print(f"{i}: {pdist(previous_point,point)}") previous_point = point i += 1 Is this the best way or is there a more pytorchic way of doing it? Continue reading...
