API#

Python API#

import_grasps_from_file

Parse an isaac_grasp YAML file for use in Isaac Sim.

GraspSpec


import_grasps_from_file(
file_path: str,
) → GraspSpec#

Parse an isaac_grasp YAML file for use in Isaac Sim. The resulting GraspSpec class will allow you to look up the data for each grasp by its name.

Parameters:

file_path (str) – A path to an isaac_grasp YAML file with format version 1.0.

Returns:

A data class that stores the information needed to replicate a grasp in Isaac Sim. This class also includes convenience functions to compute the desired pose of the gripper frame as a function of the of object position, or to compute the desired object position as a function of gripper pose.

Return type:

GraspSpec

class GraspSpec(imported_data: dict)#

Bases: object

get_grasp_names() → List[str]#

Get a list of valid grasp names stored in the imported isaac_grasp file.

Returns:

List of valid grasp names.

Return type:

List[str]

get_grasp_dict_by_name(name: str) → dict#

Get a dictionary of all data associated with a specific grasp name.

Parameters:

name (str) – Valid grasp name.

Returns:

Dictionary containing the data associated with this grasp. This includes

confidence (float):

A confidence value between 0.0 and 1.0 indicating the quality of this grasp.

position (np.array):

Translation of the gripper frame relative to the rigid body frame.

orientation (dict):

Dictionary with w and xyz components that define the orientation of the gripper frame relative to the rigid body frame.

cspace_position (dict):

A dictionary mapping each DOF that is considered to be part of the gripper to the position it was in when grasping this object.

pregrasp_cspace_position (dict):

A dictionary mapping each DOF that is considered to be part of the gripper to its open position. I.e. a grasp of this object can be acheived by moving from the pregrasp_cspace_position to cspace_position while the gripper is at the relative pose specified by position and orientation.

Return type:

dict

get_grasp_dicts() → dict#

Get a dictionary of dictionaries that specify each grasp in the imported file. The get_grasp_dict_by_name() function describes the content of each inner dictionary, and the get_grasp_names() function provides the keys to this dictionary.

Returns:

A dictionary of dictionaries that define each grasp in the imported file.

Return type:

dict

compute_gripper_pose_from_rigid_body_pose(
grasp_name: str,
rb_trans: array,
rb_quat: array,
) → Tuple[array, array]#

Given a position of the rigid body in the world or robot frame, compute the position of the gripper in that same frame to replicate the grasp associated grasp_name.

Parameters:

  • grasp_name (str) – Name of an imported grasp.

  • rb_trans (np.array) – Translation of the rigid body in the desired frame of reference.

  • rb_quat (np.array) – Quaternion orientation of the rigid body in the desired frame of reference.

Returns:

Translation and orientation of the gripper in the desired frame of reference.

Return type:

Tuple[np.array, np.array]

compute_rigid_body_pose_from_gripper_pose(
grasp_name: str,
gripper_trans: array,
gripper_quat: array,
) → Tuple[array, array]#

Given a position of the gripper in the world or robot frame, compute the position of the rigid body in that same frame to replicate the grasp associated grasp_name.

Parameters:

  • grasp_name (str) – Name of an imported grasp.

  • gripper_trans (np.array) – Translation of the gripper in the desired frame of reference.

  • gripper_quat (np.array) – Quaternion orientation of the gripper in the desired frame of reference.

Returns:

Translation and orientation of the rigid body in the desired frame of reference.

Return type:

Tuple[np.array, np.array]

On this page