Skip to content

Further updates to manifolds docs#840

Merged
dehann merged 1 commit into
JuliaRobotics:masterfrom
mateuszbaran:patch-2
Apr 7, 2022
Merged

Further updates to manifolds docs#840
dehann merged 1 commit into
JuliaRobotics:masterfrom
mateuszbaran:patch-2

Conversation

@mateuszbaran
Copy link
Copy Markdown
Contributor

I've read the page once again and changed to notation to something more similar to JuliaManifolds. Tangent vectors are denoted by capital letters X, Y, Z, Lie algebras are written using \mathfrak and manifolds using \mathcal fonts.

I have a few notes to clarify some parts:

  1. What do you mean by "Multiple X = log(M,p,q) types" in Q2? The fact that on some manifolds logarithmic map may have multiple values? That there are multiple ways to compute or approximate log (not really on (most of) the manifolds you use, but it's important for Stiefel for example)?
  2. Regarding "Projections are somewhat ambiguous, since on the one hand it might mean that a point from the manifold is projected up onto a nearby tangent space.", this is not something we do in Manifolds.jl. Projections either project a point in the embedding to a point on the manifold, or a vector from the embedding onto a tangent space at a certain point. The operation in Manifolds.jl that does more or less "a point from the manifold is projected up onto a nearby tangent space" is what we call an inverse retraction.

I've read the page once again and changed to notation to something more similar to JuliaManifolds. Tangent vectors are denoted by capital letters X, Y, Z, Lie algebras are written using `\mathfrak` and manifolds using `\mathcal` fonts.

I have a few notes to clarify some parts:

1. What do you mean by "Multiple `X = log(M,p,q)` types" in Q2? The fact that on some manifolds logarithmic map may have multiple values? That there are multiple ways to compute or approximate `log` (not really on (most of) the manifolds you use, but it's important for Stiefel for example)?
2. Regarding "Projections are somewhat ambiguous, since on the one hand it might mean that a point from the manifold is projected up onto a nearby tangent space.", this is not something we do in Manifolds.jl. Projections either project a point in the embedding to a point on the manifold, or a vector from the embedding onto a tangent space at a certain point. The operation in Manifolds.jl that does more or less "a point from the manifold is projected up onto a nearby tangent space" is what we call an inverse retraction.
@dehann
Copy link
Copy Markdown
Member

dehann commented Apr 7, 2022

Ah great thanks!


What do you mean by "Multiple X = log(M,p,q) types" in Q2?

We should probably clarify that properly here. I'm referring to the micro Lie theory discussion:

I'll send a small PR review over with the change to fix this.

Regarding "Projections are somewhat ambiguous...

Can I just copy this in? I'll send a second small PR...

@dehann dehann merged commit 75bd80d into JuliaRobotics:master Apr 7, 2022
@mateuszbaran
Copy link
Copy Markdown
Contributor Author

What do you mean by "Multiple X = log(M,p,q) types" in Q2?

We should probably clarify that properly here. I'm referring to the micro Lie theory discussion:

OK, I see. I will try to find a good wording for that -- on one hand we have different parametrizations, on another one we have different metrics.

Regarding "Projections are somewhat ambiguous...

Can I just copy this in? I'll send a second small PR...

Sure, of course!

@dehann
Copy link
Copy Markdown
Member

dehann commented Apr 7, 2022

@mateuszbaran , I was not able to request a review through the Github interface on the two small PRs,

@dehann
Copy link
Copy Markdown
Member

dehann commented Apr 7, 2022

both are there, thanks very much for taking a look!

@mateuszbaran
Copy link
Copy Markdown
Contributor Author

No problem!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

Status: No status

Development

Successfully merging this pull request may close these issues.

2 participants