Network X is a visualization and analytic platform for Python which focuses on creating and analysing network graphs. These are graphs of interconnected nodes used for representing relationships such as those found in social networks. A similar browser based network graph framework is Sigma JS, a javascript library for displaying interactive network graphs, similar to to the widely used and all powerful D3. They both excel in different areas. Network X excels in analysis and is also great for dynamically creating graph trees, while Sigma is a front end interactive GUI framework. I've been utilizing both of these lately and thought I would share some code I've written up in Python for assembling a graph network in Python using Network X and exporting the resulting JSON in a format for Sigma.
The example code below is part of a larger program that displays word associations from twitter posts. I haven't included the entire Twitter mining component as that is quite complex and uses a lot of natural language processing which isn't related to the use of either Network X or Sigma. The input (list_in) is a list of word associations (index 0 and 1), their ranking (index 2, an integer) and whether they are a primary 'P' or tertiary node 'T' (index 3). This is based around their position in relation to a central node, in this case, the work 'country'.
The output from the print statement can be dropped into any Sigma.JS canvas to be displayed and looks like this.
The example code below is part of a larger program that displays word associations from twitter posts. I haven't included the entire Twitter mining component as that is quite complex and uses a lot of natural language processing which isn't related to the use of either Network X or Sigma. The input (list_in) is a list of word associations (index 0 and 1), their ranking (index 2, an integer) and whether they are a primary 'P' or tertiary node 'T' (index 3). This is based around their position in relation to a central node, in this case, the work 'country'.
The output from the print statement can be dropped into any Sigma.JS canvas to be displayed and looks like this.
