This package contains a collection of environments for reinforcement learning with python. The environments are simulations in PhysX based on rai and the corresponding python bindings rai-python.
rai-pythonhas to be cloned into$HOME/git/and built there (use PhysX=1)
Then, clone this repo and install the project with editable files:
$ pip install -e .
To just install (without cloning the package) you can also use:
$ pip install git+ssh://git@github.com/ischubert/ryenv.git@master#egg=ryenv
A simple environment of a disk being pushed on a 2D table by a 2D finger.
Like DiskEnv, but there is a randomly generated maze on the table.
Box on a table, 10DoF
Disk on a table, with "sticky" pick-and-place