Query About Karr's Algorithm Project for GSoC 2025

117 views
Skip to first unread message

Rushabh Mehta

unread,
Jan 28, 2025, 9:34:57 AM1/28/25
to sympy
I hope this message finds you well! I previously introduced myself as someone interested in GSoC 2025, and I’ve been exploring the list of ideas on the SymPy wiki. One project that caught my attention is the implementation of Karr's algorithm for symbolic summation.

From my understanding of the codebase—particularly sympy/concrete/gosper.py—it appears that this project is still relevant for SymPy and GSoC. As far as I can tell, only Gosper's algorithm has been implemented so far, leaving a gap for Karr's algorithm.

Would it be okay to start researching and preparing for this project? what is the protocol here?

Best regards,

Rushabh

Rushabh Mehta

unread,
Feb 7, 2025, 10:22:11 AM2/7/25
to sympy

Hello again,

I wanted to follow up on my previous message regarding the implementation of Karr's algorithm for symbolic summation. I'm still very interested in contributing to this project for GSoC 2025, and I'd appreciate any clarification on whether this aligns with SymPy's current goals.

If this is indeed a relevant project, I'd love to begin preparing more thoroughly and draft a stronger proposal when the time comes. Any insights would be extremely helpful.

Thank you for your time and consideration!

Best regards,
Rushabh Mehta

Oscar Benjamin

unread,
Feb 7, 2025, 11:14:30 AM2/7/25
to sy...@googlegroups.com
Implementing Karr's algorithm definitely aligns with the SymPy roadmap
in the sense that we would want that algorithm to be implemented. It
would not be an easy project though and in fact it is likely too big
for a single GSOC project. The first steps to get to Karr's algorithm
would potentially involve improving other parts of SymPy to make sure
that all of the prerequisites are in place. It would be some work for
someone just to figure out what exactly those prerequisites are
though.

Implementing even just part of Karr's algorithm would be a great
project but it would requires at least one of:

- An unusually strong and well prepared applicant with a good
knowledge of the mathematical and computational theory.
- A supervisor who has already invested considerable time
understanding the algorithm themselves and has identified how to break
it down into manageable GSOC projects.

I don't think we have the latter so the applicant would have to be
very strong for the project to be worthwhile. It would not be worth
the time from the supervisor to familiarise themselves with the
algorithm otherwise.

Karr's paper describing the algorithm is here:
https://dl.acm.org/doi/10.1145/322248.322255

Anyone proposing a project around Karr's algorithm would need to first
understand enough of that paper to be able to say what prerequisites
it requires in terms of polynomials, rings, fields etc and whether or
not SymPy already has those things.
> --
> You received this message because you are subscribed to the Google Groups "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sympy+un...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/sympy/35bb68ba-3f0e-4752-b7c4-9b09b014aeben%40googlegroups.com.

Rushabh Mehta

unread,
Feb 7, 2025, 11:21:32 AM2/7/25
to sympy

Hello,

Thank you so much for the detailed and thoughtful response. I really appreciate it.

Based on your feedback, I think it might be more practical for me as a contributor to explore other projects that are more feasible within the scope of a single GSoC period. That being said, I'll keep this discussion on my mind and research a bit more on the  prerequisites  needed.

Thank you again for your guidance!

Best regards,
Rushabh Mehta

Reply all
Reply to author
Forward
0 new messages