This is an implementation of a Dempster-Shafer network with extensions for learning/updating the joint conditional belief matrices based on evidence observations at the nodes.

network/DSNetwork.py contains the primary Dempster-Shafer network code.
network/NetworkGenerator.py contains several test network to demonstrate how to create and use the code as well as test the functionality.
metrics contains both metrics analysis tools and plotting tools designed to facilitate understanding of the data contained within the network as well as the decisions that the network makes.