pygraph creates simple and quick Directed Graphs using relational sentences (knowledge base). It uses pydot for rendering the graphs.

The concept of this module is based on the ontological relationships, or Linked Data.

Means we don't need to create nodes then link them to one another separately, but instead we will introduce new relations between nodes through statements, let's call them relation statements.

For example, the relation statement A likes B this will be translated into a graph (fig. below) with two nodes A and B where the link (edge) between them is directed from A to B with the label likes.

New relation statements can be introduced in two ways

1. module level by using add_relation('A likes B') method, or
2. from an external file where multiple relation statements can be parsed at once.

1. module level entry example:

import sys
sys.path.append("../../pygraph")
from pygraph.dgraph import PyGraph
g = PyGraph()
g.add_relation('A likes B')
g.add_relation('A dislikes C')
g.add_relation('B loves C')
g.add_relation('C likes A')
g.draw_graph()

will produce:

2. Or you can read relations directly from an external file and pass its path as an argument for pygraph/dgraph.py.

e.g.

file name kb-person.csv and its contents:

Man is-a Person
Woman is-a Person
Child is-a Person
Father has Child
Mother has Child
Mother is-a Wife
Father is-a Husband
Husband marriedTo Wife
Wife marriedTo Husband
Husband is-a Man
Wife is-a Woman

Run:

$ python pygraph/dgraph.py example/dataset/kb-person.csv

will produce the following:

For help and options see $ python pygraph/dgraph.py -h

• pydot 1.0.2+

• Repeated relations between the same nodes are ignored.
• Output will be saved into an output folder pygraph-output.
• You may specify the delimiter in sentences (default is one space " ").
• Tested on Python 2.7.6 and 3.4.2
• To install pydot for:
  • python 3+ see James Mills branch.
  • python 2 try $ pip install pydot or $ sudo pip install pydot

• Test unit to validate the input relation sentences.
• Add more arguments to argparse options.
• Add setup.py and requirements.txt for installation.
• Extend to more graph types.