Is there a minimal / generalized coordinate version of SemiImplicitIntegrator on the roadmap or already implemented somewhere? Specifically the model used in dFlex. This would make it a lot easier to port things over.
Alternatively, any suggestions for scaling joint_act and the following parameters for the current maximal coordinate SemiImplicitIntegrator to match the dFlex implementation?
I'm setting model.joint_act in the same way, and after calling wp.sim.eval_ik, getting significantly different results for joint_qd compared to in dFlex.
Is there a minimal / generalized coordinate version of
SemiImplicitIntegrator
on the roadmap or already implemented somewhere? Specifically the model used in dFlex. This would make it a lot easier to port things over.Alternatively, any suggestions for scaling
joint_act
and the following parameters for the current maximal coordinateSemiImplicitIntegrator
to match the dFlex implementation?I'm setting
model.joint_act
in the same way, and after callingwp.sim.eval_ik
, getting significantly different results forjoint_qd
compared to in dFlex.