-
Notifications
You must be signed in to change notification settings - Fork 12
Add gudhi backend #39
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: rc-0.2.0
Are you sure you want to change the base?
Conversation
3b2ca40
to
5bf6c6a
Compare
5bf6c6a
to
471582d
Compare
55dbb23
to
984832b
Compare
Thanks for doing all of this! |
Thanks for the context @ctralie. I agree that gudhi has come a long way since then. As I do this refactor, I realize how much more mature and built out it is. I think you are correct about gudhi with respect to alpha, Cech, and maybe even extended filtrations/persistence. Maybe the point isn't to be the best computational library available but rather, pedagogical for both mathematics and python code. I really like the idea of providing an open source library useful to non-topologists but also non-programmers. I think some of gudhi's documentation is lacking and the underlying C++ documentation is doxygen based, which I'm also not a fan of. I keep thinking we should do the bare minimum here to get a working package again and then gently push people toward gudhi's implementations for the heavier things. Of course, I'm happy to modify or abandon this plan as you see fit. I also hope that you can expand on this comment:
|
I totally agree with this philosophy @catanzaromj. I would love to see a fully fledged alpha filtration with simple python libraries that we can completely explain, especially if we can fix this numerical precision bug. As for the reduction library, I was just speculating. Is there some simple reduction library that uses some of the tricks that phat does? Or perhaps we could code one ourselves? I'm not up on the current software ecosystem for tda |
This PR will add gudhi as a computational backend for now. We'll keep the code extensible for other backends and we'll try to add another in an upcoming PR.