Python libraries for robot learning, including optimization, Graph-DDP, locally weighted regression, neural networks for regression and classification, geometry and kinematics calculation, robot control interface of Baxter, PR2, Robotiq, ROS utility, etc.
Akihiko Yamaguchi, http://akihikoy.net/
CMA-ES (src/ay_py/thirdp/cma.py) is implemented by Nikolaus Hansen. Read src/ay_py/thirdp/CMA1.0.09-README.txt for more information.
- Python: core, numpy, scipy
Binaries are available (you can use apt-get):
- ROS core system, rospy, roscpp, std_msgs, std_srvs, geometry_msgs, tf, ...
- ros-ROS_DISTR-moveit-full (ROS_DISTR: groovy, hydro, indigo, etc.)
- ros-ROS_DISTR-moveit-resources
Working with Baxter/PR2 (OPTIONAL):
- Need SDK packages for Baxter/PR2
- Information would be found in: http://akihikoy.net/notes/?text
Add src/ay_py/
to Python package path. Under ay_py
, there are sub modules:
import sys
sys.path.append('YOUR_DIRECTORY/ay_py/src')
ROS-independent module, including optimization, Graph-DDP, locally weighted regression, neural networks for regression and classification, geometry and kinematics calculation, etc. You can use this module WITHOUT ROS. Just import the package:
from ay_py.core import *
Packages from third party are stored. They are internally used in ay_py.*.
ROS-dependent module, including utility, robot control interface of Baxter, PR2, Robotiq, etc. ROS environment should be setup before using this module.
from ay_py.ros import *
Useful tools, such as GUI with Python.
Requirements: for tool/py_gui.py: tmux rxvt-unicode-256color
For example, run:
python src/ay_py/tool/py_gui.py
In demo
directory, many demonstration scripts using ay_py.core
are contained. Directly run them.
In demo_ros
directory, some demonstration scripts using ay_py.ros
are contained. Directly run them.
Send e-mails to the author.