Add an argument to a specific group, creating the group if it doesn't exist. Args: group_name: Name of the group to add the argument

Load gripper state from disk. Args: config (GripperConfig): Configuration object containing necessary parameters

Format a value for display, truncating large arrays.

Register a new argument group with the global registry. Args: parser: The argparse parser to add the group to group_name: Nam

Get the current simulation app instance, initializing it if needed. Args: file_name (str): Name of the file calling this function (e.

Start Isaac Lab if it is not already running. Args: file_name (str): Name of the file that started the lab headless (bool): W

Generate a new seed that is reproducible but varied. Uses a hash of the original seed combined with a counter to ensure each

Print text in colored format and reset to default color. This function accepts all the same arguments as the built-in print function.

Apply gripper configuration if specified in args. Args: args: The parsed arguments object Returns: bool: Tr

Get the simulation app, initializing it if needed. Returns: The simulation app instance or None if initialization failed

Apply gripper configuration overrides to the args object. Only overrides values that weren't explicitly provided on the command line.

Convert a single USD file to OBJ format. Args: usd_path: Path to input USD file obj_path: Path to output OBJ file sca

Get gripper configuration based on gripper type string. Args: gripper_type (str): One of the predefined gripper types

Predict the file path that would be generated by create_isaac_grasp_data functions without actually writing any data. Args:

Convert YAML grasp format to JSON format expected by load_grasps. YAML format has grasps as a dictionary with grasp names as

Get the filepath for the debug cache.

Helper function to visualize the object mesh. Args: vis: Meshcat visualizer instance object: GuessObject instance col

Analyze a single grasp file. Args: file_path: Path to the grasp file detailed: Whether to include detailed output

Check every body except the fingers for collision against the object in transforms positions. is_invalid[i] set to Grasp.IN_COLLISION

Collect convergence-related arguments for gripper creation.

Create and return a meshcat visualizer instance.

Extract grasp names and their confidence values.

Extract mesh data from a USD file and return as a trimesh object. Args: usd_path: Path to the USD file scale: Scale fact

Extract transforms and confidences from graspgen JSON format.

Load grasp data from YAML file and return grasps dict, object file, and gripper file.

Load NPZ file and return as dictionary.

Load and scale an object mesh from file. Args: object_mesh_path: Path to the object mesh file object_scale: S

Save grasps to a YAML file. Args: grasps: Dictionary of grasps to save output_file (str): Path to save the grasps to

Shrink the GraspGuessBuffer to only contain the actual number of grasps found. Creates new arrays with the correct size and copies th

Visualize all meshes from a debug folder. Args: folder_path: Path to debug folder folder_name: Optional name for grou

Parameters ---------- origin : wp.vec3 The origin point normal : wp.vec3 The normal vector axis : int The

Given the a normalized vec3, return the U from normal.T = U*Sigma*V.T. Sigma and V.T are always 1 with a normal vector, so don't return those

Create the YAML configuration file for GraspGen.

Convert quaternion and position to 4x4 transformation matrix.

Setup PVD recording (requires Isaac Lab to be started).

Helper function to visualize normals at sample points. Args: vis: Meshcat visualizer instance points: Numpy array of shape (N

Helper function to visualize coordinate frames at transform locations. Args: vis: Meshcat visualizer instance transforms: Num

Add argument only if it doesn't already exist

Analyze confidence values between two grasp sets.

Analyze files against specified criteria. Args: files: List of file analysis results min_success: Minimum successful gra

Analyze all grasp files in a directory. Args: directory_path: Path to the directory detailed: Whether to include detaile

Check if two arrays are equal, handling different dtypes and shapes.

Load grasp data files from input folder and create balanced files in output folder. For files that contain grasp data (with 'grasps.transfor

Batch convert USD files to OBJ format based on a scales JSON file. Args: object_scales_json: Path to JSON file containing object

Calculate comprehensive statistics from grasp data. Args: confidence_values: List of confidence values success_flags: Li

Calculate validation statistics.

Check if existing grasp file exists and determine if more grasps are needed.

Combine existing grasps with new ones, filtering to only what's needed.

Compare two NPZ files and display differences.

Compare two values and return (is_equal, difference_description).

Convert absolute path to relative path from current working dir.

Convert a USD file to OBJ format using direct Python function calls. Args: usd_path: Path to the USD file scale: Scale f

Convert YAML grasp file to JSON graspgen format.

Create gripper and plot its body positions

Create a debug package containing all necessary files. Args: yaml_file: Path to the input YAML file output_zip: Path to

Create the basic grasp data structure for JSON output.

Create GraspGen configuration files for a gripper. This function generates two configuration files in the GraspGen format: 1. A YAML