Once graph data is prepared, it can be analyzed and visualized. The next goal is to get a high-level understanding of the graph:
- Is it all connected together or in many separate parts?
- Is it a hierarchy?
- Is it sparse, or is it densely connected?
- Are there obvious clusters?
These are the questions that this chapter addresses.
Statistics can provide a wealth of information, and some high-level statistics will answer questions about size, density, and number of separate graphs.
Layouts are an important visual technique to get a sense of the graph structure. Different layouts will reveal different aspects of the graph, enabling different types of analyses and supporting different types of stories.
A wide variety of node-and-link layouts can provide different ways of revealing the connections, groupings, and sequences in graphs. Other types of graph layouts focus on other properties of a graph, revealing flows, hierarchies, or multiple attributes.
Basic Graph Statistics
You can compute a wide variety of graph statistics. Which graph statistics are relevant depends, in part, on your objective. Some of the simpler graph statistics are outlined here, such as density, degree, and centrality.
Size (Number of Nodes and Number of Edges)
As described in Chapter 3, “Data—Collect, Clean, and Connect,” you can collect various simple graph statistics (such as graph size statistics—that is, number of nodes and number of edges) during the data preparation. ...