How to get the dtw/*pair_info.json?

[BigJohnn]

Hi, thanks for the great work!

I'm trying to reproduce the DTW-based preprocessing pipeline, but I'm not sure how the dtw/*pair_info.json files are generated. I couldn't find scripts or documentation describing how to produce these JSON files.

Could you clarify the following?

How are the pair_info.json files created?

Is there a specific script or command for generating them?

If possible, could you provide an example or minimal code snippet showing how to generate the DTW metadata from raw trajectories?

This would be extremely helpful for reproducing your results.
Thanks a lot for your time!

[Xunmenggod]

Hi, I would also like to replicate this wonderful project. But I met with the same problem that I do not know the exact way to build the dtw pair of two datasets. If you could open-source the implementation of dtw, I would appreciate it a lot!