Hi, thanks for the valuable dataset contribution.
I use the Agilex_Cobot_Magic_close_drawer_bottom dataset as an example.
dataset = LeRobotDataset(repo_id="Agilex_Cobot_Magic_close_drawer_bottom")
item = dataset[200]
print(item['eef_sim_pose_state'])
# tensor([ 0.2329, 0.2941, 1.0534, 2.0357, -0.0219, 1.5082, 0.5111, -0.0441, 0.8445, 1.9425, -0.0792, 1.9126])
print(item['observation.state'][7:10]) # corresponding to left_eef_pos_x_m and y and z
# tensor([-0.0059, -0.0073, 0.2785])
Why the eef_sim_pose_state is different from the left_eef_pose_x_m? Is it because that the definitions of the base of coordination system are diverse, as you mention in readme?
Hi, thanks for the valuable dataset contribution.
I use the
Agilex_Cobot_Magic_close_drawer_bottomdataset as an example.Why the eef_sim_pose_state is different from the left_eef_pose_x_m? Is it because that the definitions of the base of coordination system are diverse, as you mention in readme?