Efficient Equations of Motion Generation

[LUIS ANGEL MELENDEZ CHAVERO]

Hello everyone,

I'm applying for GSoC 2025 with SymPy and would like to share my draft proposal for feedback.

Project: Efficient Generation of Equations of Motion
GitHub Notebook: https://github.com/LuisOfL/GSoC2025-SymPy/blob/main/GSoC.ipynb
Pull Request:  https://github.com/sympy/sympy/pull/27858

I welcome any feedback to improve my proposal before I officially submit it.

Thanks in advance!

[Jason Moore]

Dear Luis,

Thank you for your interest. I scanned your PR and proposal. Unfortunately, these ideas are not well aligned with the needs of the SymPy Mechanics package. I suggest doing more research on what the issues are regarding efficient equation generation and reworking your proposal. To improve the generation you would need to profile the code in slow examples and find ways to speed them up. Using simplification routines will most likely always slow them down.

Jason

moorepants.info
+01 530-601-9791

--
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/4ac5826c-841a-4e41-9751-6f7828aa2ad3n%40googlegroups.com.

[LUIS ANGEL MELENDEZ CHAVERO]

Hi,

I hope you're doing well.

I just wanted to follow up on my GSoC 2025 proposal for the SymPy project (Efficient Equations of Motion Generation). Based on your previous feedback, I’ve made some improvements to both the proposal and the contribution itself.

I ran some profiling to identify bottlenecks, made a focused optimization in lagrange.py, and also put together a Jupyter notebook with the results.

Here’s the updated patch:
🔗 https://github.com/sympy/sympy/pull/27870

If you have a moment to take another look, I’d really appreciate any additional thoughts or suggestions—especially if there’s anything else I should refine or explore further.

Thanks again for your time and support!

Best,
Luis

[peter.st...@gmail.com]

Dear Luis,

 

likely my python / sympy knowledge is not good enough, but I do not see any difference in 27870 between what was and what you did.

Yet you get 25% speed improvement.

 

Best regards,

 

Peter

To view this discussion visit https://groups.google.com/d/msgid/sympy/a4599617-f8e8-471d-9990-8015dbad05f8n%40googlegroups.com.

[LUIS ANGEL MELENDEZ CHAVERO]

Hi Peter,

Thank you very much for your reply!

To be honest, I'm still learning Python and SymPy. When I generated a cProfile report, I got the following results: https://github.com/LuisOfL/GSoC2025-SymPy/blob/main/cProfile.txt. It’s worth noting that I may have misunderstood the 25% — it refers to a specific function call, not the overall class.

In the link I shared with you, the line
1 0.000 0.000 0.380 0.380 lagrange.py:174(form_lagranges_equations)
was reduced from ~0.380 s to ~0.285 s in the cProfile output.

[peter.st...@gmail.com]

Dear Luis,

 

I did not doubt your statement.

Just due to my limited python knowledge I simply cannot see what you changed.

 

Good luck!

To view this discussion visit https://groups.google.com/d/msgid/sympy/532a2586-3074-4ffe-beac-044904cfb7e0n%40googlegroups.com.