Open Daniel63656 opened 2 months ago
Unfortunately, I don't have plans to update this package in the near future, as I'm very overloaded and the current python implementation is fast enough for my usecases.
Regarding using C to speed up computations, it would be great indeed. When I made this port, the original APTED code was in Java, so I couldn't do it, but now, the group that proposed the algorithm has a C++ implementation: https://github.com/DatabaseGroup/tree-similarity/
Maybe a wrapper around that would be feasible. If you (or anyone else) do that in a separate package, let me know (or submit a pull request), so I can point your solution in the README.
Great package that is already the fastest python implementation as far as I know. Still, are there plans to use C code to speed up computations, similar to how Levenstheins python port does? Would make this package the definitive go to for TED.