Yes, You Can Build Network Diagrams in Tableau
To find out more about Concentra's bespoke Analytics Services and Solutions see Services.
People tend to be visually satisfied when they see hierarchical structures represented with network diagrams. Network diagrams immediately provide a clear and explanatory description of the scenario analyzed.
However, they are not a built-in option in Tableau. That may intimidate you from the start, but don't panic; you will be able to step up your dashboard game with just a short explanation of the data structure required to build these type of graphs.
Drawing networks, in general, is not conceptually hard - at least when it comes to doing it in Tableau! Essentially, you will have to draw your network and instruct Tableau on how it should connect your nodes.
For the purpose of this blog, I decided to analyze British Airways global connections from London Heathrow Airport. I scraped some basic data from Heathrow Airport's website and processed it with Alteryx to get the data in the correct format.
Alteryx is a great tool for data preparation, allowing you to clean your data gradually by constructing visually intuitive workflows (Figure 2). SQL or Excel are valid alternatives if you are not familiar with it.
As mentioned earlier, the key concept is to instruct Tableau on how to connect the network nodes. Each point will need a set of defined properties:
- A pair of coordinates (X-Y, lat-lon): If you are using geographical data that Tableau "understands", then you can bypass this as latitude and longitude will be automatically generated
- Path order: This is essential in determining the order in which the nodes are connected to each other
- Path ID: Used to identify each connection. It can be useful for reference or general formatting when building the dashboard
Once these properties have been defined, we can use Tableau's formatting options to add value to the visualization. You can highlight groups of routes by continent and see how many destinations there are within a country, like I did in the visualisation below.