search

jstmn / cppflow

Open source implementation of "CppFlow: Generative Inverse Kinematics for Efficient and Robust Cartesian Path Planning" (ICRA 2024)
https://jstmn.github.io/cppflow-website/
10 stars 0 forks source link

Why doesn't the damping factor（opt_params.lm_lambda） get updated in your cpp-flow code when optimizing trajectories using the Levenberg-Marquardt algorithm? #1

Open longhouyunhhh opened 1 month ago

longhouyunhhh commented 1 month ago

Why doesn't the damping factor（opt_params.lm_lambda） get updated in your cpp-flow code when optimizing trajectories using the Levenberg-Marquardt algorithm?

jstmn commented 1 month ago

Sorry what do you mean exactly? It's used here: https://github.com/jstmn/cppflow/blob/master/cppflow/optimization.py#L158 if you're suggesting it's not used. Or are you saying why isn't it updated throughout the optimization?