rofunc.learning.RofuncRL.tasks.isaacgymenv.shadow_hand#
1. Module Contents#
1.1. Classes#
1.2. Functions#
1.3. API#
- class rofunc.learning.RofuncRL.tasks.isaacgymenv.shadow_hand.ShadowHand(cfg, rl_device, sim_device, graphics_device_id, headless, virtual_screen_capture, force_render)#
Bases:
rofunc.learning.RofuncRL.tasks.isaacgymenv.base.vec_task.VecTask- create_sim()#
- compute_reward(actions)#
- compute_observations()#
- compute_fingertip_observations(no_vel=False)#
- compute_full_observations(no_vel=False)#
- compute_full_state(asymm_obs=False)#
- reset_target_pose(env_ids, apply_reset=False)#
- reset_idx(env_ids, goal_env_ids)#
- pre_physics_step(actions)#
- post_physics_step()#
- rofunc.learning.RofuncRL.tasks.isaacgymenv.shadow_hand.compute_hand_reward(rew_buf, reset_buf, reset_goal_buf, progress_buf, successes, consecutive_successes, max_episode_length: float, object_pos, object_rot, target_pos, target_rot, dist_reward_scale: float, rot_reward_scale: float, rot_eps: float, actions, action_penalty_scale: float, success_tolerance: float, reach_goal_bonus: float, fall_dist: float, fall_penalty: float, max_consecutive_successes: int, av_factor: float, ignore_z_rot: bool)#
- rofunc.learning.RofuncRL.tasks.isaacgymenv.shadow_hand.randomize_rotation(rand0, rand1, x_unit_tensor, y_unit_tensor)#
- rofunc.learning.RofuncRL.tasks.isaacgymenv.shadow_hand.randomize_rotation_pen(rand0, rand1, max_angle, x_unit_tensor, y_unit_tensor, z_unit_tensor)#