Skip down to main content

A simple trick for drawing pseudo-geographical networks

Published on
21 Jul 2016
Written by
Private: Dr Graham McNeill

Not all networks have geographic information (e.g., friendship networks, hyperlink networks, etc.). However, for networks that do have geographic data (such as networks of commuting or migration), geography can be a powerful way to position the nodes in a visualization. Geographic position allows viewers to interpret the network in a geographical context and also, enables viewers that are familiar with the geography to quickly find specific nodes. However, many networks include dense clusters (Figure 1a) that make a simple geographic layout inappropriate for static images. Even for interactive visualizations, users will have to pan and zoom frequently and hence, may struggle to see the big picture properties of the network.

Figure 1: UK local authorities. (a) centroids; (b) MDS λ=100; (c) MDS λ=200.

Figure 1: UK local authorities. (a) centroids; (b) MDS λ=100; (c) MDS λ=200.

We would like to retain the advantages of a geographic layout, but avoid the clustering issue. We first started by looking at ways to modify standard network force-directed layout algorithms to achieve this, but we have found that multidimensional scaling (MDS)—a technique typically used for data visualization and dimensionality reduction—works well. Given a matrix of pairwise distances, MDS produces a set of points whose pairwise Euclidean distances are as close as possible to those in the original matrix. In our case, we compute the Euclidean distance matrix between nodes, add a number λ to every entry (except those on the diagonal) and then compute new coordinates for the nodes using MDS. Note that by adding λ to every distance, we increases the relative distance between pairs of points that are close together.

There are many variants of MDS. We use the well-known Sammon mapping which focuses on the distances between nearby points. Figure 1 demonstrates the approach for different values of λ on a map of local authorities in the United Kingdom. In this case, a value of λ between 100-200km produces reasonable results—the clusters are spread out, but significant geographic information is retained.

Related Topics:

Privacy Overview
Oxford Internet Institute

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.

Strictly Necessary Cookies
  • moove_gdrp_popup -  a cookie that saves your preferences for cookie settings. Without this cookie, the screen offering you cookie options will appear on every page you visit.

This cookie remains on your computer for 365 days, but you can adjust your preferences at any time by clicking on the "Cookie settings" link in the website footer.

Please note that if you visit the Oxford University website, any cookies you accept there will appear on our site here too, this being a subdomain. To control them, you must change your cookie preferences on the main University website.

Google Analytics

This website uses Google Tags and Google Analytics to collect anonymised information such as the number of visitors to the site, and the most popular pages. Keeping these cookies enabled helps the OII improve our website.

Enabling this option will allow cookies from:

  • Google Analytics - tracking visits to the ox.ac.uk and oii.ox.ac.uk domains

These cookies will remain on your website for 365 days, but you can edit your cookie preferences at any time via the "Cookie Settings" button in the website footer.