We can set model limits (e.g. joint position limits) using enforce_model_limits, however it would be good to be able to specify a fraction/percentage so that the limits of the optimization can be inside the limits of the robot. E.g.
enforce_model_limits(name, safe_frac=0.95) # by default safe_frac=1.0
Then the limits that are actually set in the optimization are
0.95q^- \leq q \leq 0.95q^+
Just note, be sure to handle the case when $\exists i : q^+_i \neq -q^-_i$.
We can set model limits (e.g. joint position limits) using
enforce_model_limits, however it would be good to be able to specify a fraction/percentage so that the limits of the optimization can be inside the limits of the robot. E.g.Then the limits that are actually set in the optimization are
Just note, be sure to handle the case when $\exists i : q^+_i \neq -q^-_i$.