Network chart

Network diagram with the NetworkX library

NetworkX is a Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks..

The examples below will guide you through a migration dataset already discussed in data-to-viz.com. It starts by describing the input dataset and the basic usage of the Chord() function.

Most basic network chart with Python and NetworkX

Custom network appearance: color, shape, size, links

Control the layout used for the node location

Manage directed and undirected networks by adding arrows

Control the background color of a network chart

Bundles network edges together to reduce the figure visual clutter

A common need when dealing with network charts is to map a numeric or categorical variable to the nodes or edges appearance. This is totally doable and adds some insight to the figure.

Map a continuous or categoric variable to nodes

Map a color to edges

Network chart from similarity matrix

Another commin task is to build a network chart from a correlation matrix. Let's say you have several numeric variables describing some items in a dataset. You can compute a similarity matrix and display it as a network chart. This process is described in the post #327 below.

Network chart from correlation matrix with Python and NetworkX

Network diagram with graph-tool

The graph tool library is a python library implemented in C++. It makes it highly efficient to draw networks containing many nodes. Graphics to come.