A lightweight Python3 library for visualizing high-dimensional data. It is designed for experimental research on Navigation problems. However, the same APIs could be useful for a variety of applications.
It is built on Matplotlib with the goal provide an easy interface for 2D representation and annotation of high dimensional state spaces.
For instance, using just one line of code, this library could
- Render n x n state space.
- Label each state.
- Enable state/feature level grid.
- Draw trajectories
- Visualize features such as one-hot indicator variable, vector, and RGB image.
To install current release with pip:
pip install navigation-vis
To install from source:
python setup.py install
If you're curious what it can do, check out some examples on
