It’s hard to tell what is going on. But even using linear_approximation fails, indicating a problem in the Newton step. Have you tried whether a simpler version works?
That being said, using lmmcp seems to solve the simulation.
It’s hard to tell what is going on. But even using linear_approximation fails, indicating a problem in the Newton step. Have you tried whether a simpler version works?
That being said, using lmmcp seems to solve the simulation.