graphskeleton

This module provides tools for creating and using graph skeletons for Bayesian networks. A graph skeleton in this case is a vertex set and a directed edge set, with no further information about the specific nodes.

class libpgm.graphskeleton.GraphSkeleton

This class represents a graph skeleton, meaning a vertex set and a directed edge set. It contains the attributes V and E, and the methods load, getparents, getchildren, and toporder.

V = None

A list of names of vertices.

E = None

A list of [origin, destination] pairs of vertices that constitute edges.

alldata = None

(Inherited from dictionary) A variable that stores a key-indexable dictionary once it is loaded from a file.

load(path)

Load the graph skeleton from a text file located at path.

Text file must be a plaintext .txt file with a JSON-style representation of a dict. Dict must contain the top-level keys “V” and “E” with the following formats:

{
    'V': ['<vertex_name_1>', ... , '<vertex_name_n'],
    'E': [['vertex_of_origin', 'vertex_of_destination'], ... ]
}

Arguments:

1. path – The path to the file containing input data (e.g., “mydictionary.txt”).

Attributes modified:

1. V – The set of vertices.
2. E – The set of edges.

getparents(vertex)

Return the parents of vertex in the graph skeleton.

Arguments:

1. vertex – The name of the vertex whose parents the function finds.

Returns:

A list containing the names of the parents of the vertex.

getchildren(vertex)

Return the children of vertex in the graph skeleton.

Arguments:

1. vertex – The name of the vertex whose children the function finds.

Returns:

A list containing the names of the children of the vertex.

toporder()

Modify the vertices of the graph skeleton such that they are in topological order.

A topological order is an order of vertices such that if there is an edge from u to v, u appears before v in the ordering. It works only for directed ayclic graphs.

Attributes modified:

1. V – The names of the vertices are put in topological order.

The function also checks for cycles in the graph, and returns an error if one is found.