News for hackers. Stuff that matters.

A comment on Regularized Newton Method with Global $O(1/k^2)$ Convergence. Parent comment

Re:Regularized Newton Method with Global $O(1/k^2)$ Convergence by gugagore on Wednesday December 08, 2021 @01:08PM (#29488202)
There are probably many ways to formulate the same geometric constraint. For example, what parameterization of rotations do you use? I think the choice has implications for how easy it is to solve the equations.
[ Parent | on HN ]
Re:Regularized Newton Method with Global $O(1/k^2)$ Convergence by phkahler on Thursday December 09, 2021 @09:37AM (#29497772)
>> For example, what parameterization of rotations do you use?

For orientation we use quaternions. For other things it's an axis-angle representation. For the equal angle constraint I'm not sure how that's implemented. There was a recent addition of length-ratio between lines and arcs. That implementation was surprising to me.

[ Parent | on HN ]