Visualization of Graphs and Networks-Part II

As we discussed the basics of graphs and networks in the previous tutorial, that is why graphs are important for relational data, and some examples of graphs. Along with that, we also discussed some considerations and problems that are to be kept in mind whilst drawing the graphs. In this tutorial, we’ll further dive into it, discussing different layouts, node-link diagrams and Attribute-driven layouts.

Possible Ways of Graph Visualization

Networks and Trees can be arranged in several different ways, and their choice depends on the relation, task and scale. Some arrangements are node-link diagrams, which are valid for both networks and trees, adjacency matrix, which is a derived table valid for both networks and trees, and enclosure which has containment marks and is valid for only trees.

Node-Link Diagrams

Node-link diagrams have either stable or unstable layouts.

A list of stable layouts to randomness is:

• Grid Layouts: The nodes are generally placed in a grid form, typically in rows and columns. An easy-to-understand layout and is suitable for visualizing structures with clear organization.
• Orthogonal Layouts: The nodes are placed in grid form and generally follow the 90-degree path (orthogonal). This layout is often used to represent hierarchical structures or to emphasize the flow of information.
• Circular Layouts: The nodes are generally placed in circular paths, with edges radiating from the centre. A layout useful for emphasizing relationships between nodes is often applied in cyclic structures.
• Arc Diagrams: The nodes are generally placed on a single axis and the arcs connect the related nodes. Useful in visualizing relationships in linear or sequential data.
• Chord diagram: The nodes are generally placed in a circle, and chords connect related nodes. Commonly used to show relationships between entities, such as connections between different categories or groups.
• Radial Graph Diagram: Nodes are generally placed in a circular pattern, just like a circular layout, but the edges do not necessarily follow straight lines. Radial layouts are often employed for emphasizing relationships and reducing visual clutter.
• Sugiyama Layout (Special Directed Acyclic Graph): Specifically designed for directed acyclic graphs (DAGs), Sugiyama layout generally places nodes in layers, minimizing edge crossings. It is useful for visualizing processes or hierarchies with a clear direction.

Unstable Layouts include:

• Random Layout: Nodes are generally placed in a random order. Despite it being quick and easy, it can result in confusing visualizations without any clear pattern or organization.
• Force Directed Layout: Nodes are initially placed randomly and then repel each other or are attracted based on physical simulation principles. This dynamic layout adjusts over time to minimize energy in the system, providing flexibility for visualizing complex relationships but may result in less stable arrangements.

Below are some visual examples of the above-mentioned diagrams:

Attribute-Driven Layouts

Large node-link layouts at times get messy, which can hinder the process of good visualization. To cope with such type of situation we can use data attributes to perform layouts. Attributes may be associated with nodes or edges or maybe statistical properties of the graph.

A few examples of Attribute-Driven Layouts include Skitter Layout, which shows the Internet connectivity in the form of a radial scatterplot, in which the angle shows the latitude and the radius is the degree of the number of connections.

Another example is the Hive plot, in which the nodes or dots may be replicated. In this, nodes are sorted on radial axes by network statistics. Different axes may show different subsets of nodes.

Conclusion

Conclusively, we have discussed the node-link diagrams in some detail and got an overview of the attribute-driven layouts, and why they are used. We also discussed different layouts, which will be applied in different types of visualization. In the next tutorial, we’ll discuss the Sugiyama Layout, what it is the steps are drawing a Sugiyama Layout and what is its purpose. A step-wise explanation will be provided in the upcoming tutorial.

I hope you liked the tutorial, please provide your feedback in the comments section below, or email us at immadshahid@gmail.com.