kinjax is a Python package that provides functions for computing the forward kinematics and Jacobian of a robot. It uses JAX to enable fast computations on GPUs.
To install the module, run:
pip install kinjax
To use kinjax, you need to provide it with a URDF file that describes the robot you want to compute the forward kinematics and Jacobian for. You can then call get_fk_fn and get_jacobian_fn to create functions that can compute the forward kinematics and Jacobian of the robot, respectively.
Here's an example of how to use kinjax to compute the forward kinematics and Jacobian of a robot:
import jax
import jax.numpy as jnp
import kinjax
end_effector_link_name = "hand"
dof = 6
# Load the URDF file
link_dict, joint_dict = kinjax.get_link_joint_dict("robot.urdf")
# Compute the forward kinematics function for the robot
fk_fn = kinjax.get_fk_fn(link_dict, joint_dict, dof, end_effector_link_name)
# Compute the Jacobian function for the robot
jac_fn = kinjax.get_jacobian_fn(link_dict, joint_dict, dof, end_effector_link_name)
# Compute the forward kinematics and Jacobian for a given joint configuration
q = jnp.array([0.1, 0.2, 0.3, 0.4, 0.5, 0.6])
T_fk = fk_fn(q)
J = jac_fn(q) This project is licensed under the MIT License