Skip to content

Latest commit

 

History

History
14 lines (11 loc) · 748 Bytes

README.md

File metadata and controls

14 lines (11 loc) · 748 Bytes

Here are some code about the Quasi-metric method. Thi is a very experimental (and messy) code for research purpose. For more details about the method see: http://dx.doi.org/10.1371/journal.pone.0083411

This package depends on 'graph-tool' http://graph-tool.skewed.de/ Unfortunately it seems that the pip install of this package doesn't work.

Principles of use:

  • Define your probabilistic variables.
  • Define your transition distribution
  • Build the Quasimetric towards a goal and get the Policy.

You can either provide a function to build the transition probability distribution in order to compute everything offline (see samples/test_qm.py). Or you can build everything online through successive observations (see samples/test_onlineqm.py).