search

pearl-robot-lab / rlmmbp

Learning mobile manipulation behaviors through reinforcement learning
Other
48 stars 4 forks source link

Some problems #1

Open taylorlulu opened 6 months ago

taylorlulu commented 6 months ago

I found a problem with the code while I'm debugging.

https://github.com/iROSA-lab/rlmmbp/blob/341cf0c24b64aa0329cd4b217f8fbc0d080a584a/learned_robot_placement/tasks/utils/scene_utils.py#L226

which may cause mismatched indexing. I found it more appropriate to add del objects_dimensions[goal_obj_index].

del object_positions[goal_obj_index], object_yaws[goal_obj_index]
del objects_dimensions[goal_obj_index]

If there's any misunderstanding on my part, please correct me.

sjauhri commented 6 months ago

Hi @taylorlulu,

Thanks for pointing out this fix. Feel free to submit a pull request and I will merge it