I recently happily noticed that there is some increased activity around Manifolds in the Julia related social areas.
So I want to try to maybe move some of these discussions to a digital meeting. Everyone is welcome to join!
The idea is to discuss
What’s new around Manifolds.jl and other JuliaManifolds packages?
What are participants struggle with / need help with / want to discuss?
What’s next? What’s your plans to contribute?
And for a first meeting or anyone joining anew also maybe a short introduction, if they want to – for a first meeting this then concerns everyone who wants to join.
As a first meeting time, I propose
Tuesday, September 24, 17.00 (or 5 pm) CEST on zoom
where my “What’s next?” topic will be Lie groups.
and will post a link soon (in case we move it a bit after feedback).
No I did not take notes, since it was really just an informal discussion.
We discussed a bit all different exps there are, and which ones will be the main ones in Lie groups. For that, we work on GitHub - JuliaManifolds/LieGroups.jl: LieGroups.jl – WIP, slowly rethinking redoing the interface from the GrouManifolds as they were called when introduced in Manifolds.jl. The name is also still work-in-progress (see issue 1), and overall this total rework will still take quite a while.
But is was really nice and as you see, we already have a tentative next community call to join.
Thanks for the summary. I would love to understand if there are different kinds of exp for the Lie Groups encountered in robotics (e.g. SO(3)). It is likely to affect my efforts toward developing an interface language for robotics, one that is built upon a kernel of mathematical understanding of the data. Kind of reminiscent of “symbolic reasoning”.
I’m looking for clarity on the “action spaces of robots”, and concretely I’m curious if the machinery and mathematical knowledge in the Manifolds Julia ecosystem could be helpful in a systematic and programmatic naming scheme for the different action spaces of a robot.
Robotics is one of the reason I decided to even start the LieGroups.jl this year, though I can only work on that very much on the side. There were just too many things asked on Manifolds.
Yes, that difference in exps will also be described in more details in the Lie group docs as soon as I find time. But there is other people using Manifolds.jl for robotics.
I am not sure what you are referring to with “symbolic reasoning” nor with “action spaces of robots” but maybe that is also another topic than the announcement channel for JuliaManifolds Community calls should cover.
Let’s already announce the next one – we will do another JulliaManifoldsCommunity Call on
Tuesday, November 19, 2024, 17.00 CET (UTC+1)
The zoom link will of course still stay the same. Feel free to post potential topics we could discuss. One that came up today is to discuss adjoints on Lie groups.
@kellertuer Despite my best efforts, I’m unable to attend today due to the new schedule and timing. However, as mentioned during the October 22nd meetup, I still hope to give a brief 5-minute presentation. Please extend my apologies to your colleagues so that no one feels put off. Thank you for your understanding.