Extended networkx Tools

Python Package for for visualizing and converting networkx graphs.

Introduction

This package was created for the purpose of examining bidirectional graphs with respect to its convergence rate and edge costs.

Installation

pip install extended-networkx-tools

Documentation

extended-networkx-tools.readthedocs.io

The package

Currently the package contains 3 main modules, Creator, Analytics, Solver, Visual and AnalyticsGraph.

Creator

Contains tools to create networkx graphs based on given parameters, such as randomly create an empty graph based on a number of nodes, or specify precisely the coordinates of nodes and the edges between them.

Analytics

Has tools for analysing the networkx object and extract useful information from it, such as convergence rate, neighbour matrix, its eigenvalues.

Solver

Creates greedy solutions to a connected graph taken from graph theory. The current approaches are:

• path: Adds edges as a path from the start to end node
• cycle: Adds edges just like the path, but also one edge from the start to end node.
• complete: Adds edges between all nodes to all the other nodes, such as the maximum distance between every node is one.

Visual

Is used to print a networkx graph to the screen, with its edges.

Example output graph

AnalyticsGraph

The AnalyticsGraph class is a helper class that serves the purpose of a wrapper object that can do all calculations based on changes done to the graph, rather than recalculating every metric after simple changes. Such as the connectivity state will stay the same after adding an edge.

There is also options to revert changes and keep previous calculations.

Example usage:

from extended_networkx_tools import Creator, Solver, AnalyticsGraph

# Create a random graph with a path
g = Creator.from_random(10)
g = Solver.path(g)

# Convert the graph to an AnalytcsGraph object
ag = AnalyticsGraph(g)

convergence_rate = ag.get_convergence_rate() # Calculates the convergence rate from scratch
ag.remove_edge(4, 5)                         # Removes an edge
ag.revert()                                  # Revert the changes
convergence_rate = ag.get_convergence_rate() # Doesn't calculate it since it's saved from previous state

ag.is_connected()  # Yields True. Uses BFS to check if the graph is connected into one component.
ag.add_edge(1, 4)  # Adds a random edge between 2 nodes
ag.is_connected()  # Immediately returns True, as the connectivity isn't affected by 
                   # adding an edge if it already was True before adding it.

Usage

Import

from extended_networkx_tools import Creator, Analytics, Visual, Solver, AnalyticsGraph