Actions/Joints mapping

[ghost]

Hello,
I'm trying to figure out how to map actions to joints. In other words, I need to understand which joint corresponds to each index in actions[]. Actually, I trained a policy for my quadruped robot and use the .onnx model to control it from another simulation environment (Webots) prior deploying it on the real robot. Currently, I'm mapping joints to actions based on the order they appear in the URDF file, but this isn't working. According to what I've found on forums, PhysX seems to manage joint order for optimization purposes.
Ideally, I'd like a method in play.py that displays the list of joints in the order they are actually mapped to the actions[] array..

Thanks!
Regards.

[ghost]

Ok, I found a solution here: https://forums.developer.nvidia.com/t/articulationcontroller-questions/268920
Regards.