VMAS is a vectorized differentiable simulator designed for efficient Multi-Agent Reinforcement Learning benchmarking. It is comprised of a vectorized 2D physics engine written in PyTorch and a set of challenging multi-robot scenarios. Additional scenarios can be implemented through a simple and modular interface.
Description:
I've noticed that in the navigation scenario, the x_semidim and y_semidim parameters cannot be set or utilized. It seems that the scenario is designed for navigation in "infinite" environments, which limits the ability to configure these parameters for tasks that require a bounded environment.
This limitation restricts the flexibility of the navigation scenario, as users cannot define the dimensions of the environment. If someone wishes to use the scenario in a "limited" environment, they currently have no elegant way to configure it.
Questions:
Would it be possible to to allow the use of the x_semidim and y_semidim parameters for defining a bounded environment?
Are there any specific limitations or something I'm ignoring regarding this task?
Suggestion:
Use kwargs to provide access to these two parameters.
Description: I've noticed that in the navigation scenario, the x_semidim and y_semidim parameters cannot be set or utilized. It seems that the scenario is designed for navigation in "infinite" environments, which limits the ability to configure these parameters for tasks that require a bounded environment.
This limitation restricts the flexibility of the navigation scenario, as users cannot define the dimensions of the environment. If someone wishes to use the scenario in a "limited" environment, they currently have no elegant way to configure it.
Questions:
Suggestion: Use kwargs to provide access to these two parameters.