lakehanne/gps repository overview

⁠Guided Policy Search

This code is a re-implementation of the guided policy search algorithm and LQG-based trajectory optimization, meant to help others understand, reuse, and build upon existing work. It includes a complete robot controller and sensor interface for the PR2 robot via ROS, and an interface for simulated agents in Box2D and MuJoCo. Source code is available on GitHub. This my the catkin interface⁠ to the code domiciled on Chelsea Finn's github account⁠ .

While the core functionality is fully implemented and tested, the codebase is a work in progress. See the FAQ⁠ for information on planned future additions to the code.

⁠Running

In your OS host running unix for example, we need to bind our xhost to the docker engine. Run

     xhost +local:root   

Then run this image:

docker run -it --rm -v \
                     /tmp/.X11-unix:/tmp/.X11-unix:ro \
                    -e DISPLAY=$DISPLAY \
                   lakehanne/gps:gps-v0.0.x

where v0.0.x would be the latest tag. for example v0.1.0

⁠Setup

Do the following in terminal

     cd /root/catkin_ws

⁠Setup python path and ros environment

     bash setup.sh

⁠Examples

In order to run the examples, we need to enable the QT/KDE environment.

Add this line QT_X11_NO_MITSHM=1 in ~/.bashrc or export the environmental variable from terminal:

export QT_X11_NO_MITSHM=1

⁠Box Examples

To run the box point mass example, do

     bash run_box.sh

⁠PR2 ROS Examples

Launch gazebo and the GPSPR2Plugin:

   roslaunch gps_agent_pkg pr2_gazebo.launch

1. Run the following from the gps directory:

   python python/gps/gps_main.py pr2_example

For other setup, please see the gps page at http://rll.berkeley.edu/gps/⁠