Another Tree for the Forest
The Kruskal algorithm, which constructs a Minimum Spanning Tree (MST) diagram, forms the foundation of the collaboration between cosmologist Krishna Naidoo and designer Ed Cornish. Their process draws on Krishna’s PhD research, which uses Minimum Spanning Tree diagrams to measure the distances between galaxies in order to extract information hidden within selected data sets.
Whilst algorithms are commonly spoken of in almost mystical ways, they are simply a set of instructions or rules. Ed & Krishna’s collaboration aims to demystify algorithms by visualising their processes and communicating their inherent value in collecting, collating and creating information.
Ed & Krishna began their collaboration by testing the Kruskal algorithm at much smaller scales than Krishna had been using in his galaxy research. The first test MST they created was an animated map of the world, using the algorithm to create links between airports.
The First Test – creating MST from world airport locations
From this two dimensional map, the team began to explore the Kruskal algorithm in three-dimensional space, and it quickly became apparent that the 2D graphs and diagrams generated did not clearly communicate the three-dimensional nature of the data. Pursuing this idea, Krishna & Ed created a 3D test model of a Minimum Spanning Tree based on 100 randomly generated points. The intention was to simulate on a small-scale what a 3D model of the galaxy data might look like.
3D Test Model created from random data points.
Once they had completed their initial test models, Ed & Krishna focused on the workings of the Kruskal algorithm itself. Algorithms surround us in the modern world, but their inner workings are a mystery to many. The team’s intention was to expose how algorithms work and to demystify their processes. Krishna produced a series of GIF animations of the algorithm slowly constructing a Minimum Spanning Tree from a selection of random data points. The team found watching the algorithm’s process to be strangely captivating – it exposed the methodical and logical nature of the algorithm, steadily measuring every possible connection and constructing the MST without any room for human error. From this they developed their immersive fulldome film, Another Tree for the Forest, showing the algorithm at work creating a Minimum Spanning Tree (MST) using a dataset of the 500 brightest stars above Grizedale Forest at the time of the final exhibition.
Still images from ‘Another Tree for the Forest’ fulldome film.
The 2D appearance of the stars above us belies the three-dimensional nature of the distances between them, Ed & Krishna’s film uses the Kruskal algorithm to take the audience on a journey through the stars and creates a new tree in space above Grizedale Forest. Beginning with the familiar 2D earth-bound perspective the moves outward to reveal the stars in an expanded three-dimensional MST structure which allows visitors to see the stars above Grizedale from new perspectives. The accompanying soundtrack is also made by the algorithm using the same the star data – distances between stars are given specific notes, and played on different instruments depending on whether the distance measured forms part of the final MST.
The film and soundtrack are created in collaboration with the algorithm – Krishna & Ed hope to communicate that whilst scientists may use algorithms in their work, human interaction is still necessary to ‘prune’ the data. The film is an example of how algorithms and computer visualisation can offer new perspectives on familiar data, but also evidences the need for human interaction in interpreting this information.
Additional artworks accompanying the Another Tree for the Forest film in the final exhibition include the initial 3D test model of the Minimum Spanning Tree diagram, an algorithmic music box and the interactive Algorithmic Tree Wall installation also devised by Ed & Krishna.
Ed & Krishna’s work-in-progress as presented in the Grizedale Forest exhibition.