Exploring multivariate data with the grand tour1 is a visually exciting way to discover interesting structure. However, one criticism of this method is that as dimensionality increases the chances of quickly discovering views of interest diminish rapidly, because of the random nature of the grand tour, and the expanding volume of space. To improve the chances of discovering interesting structure we propose a method for controlling the exploration by motion control and directing movement along the gradient of a projection pursuit function. The benefits of this approach are two-fold. Firstly, it provides a fast, powerful exploratory data analysis tool, and secondly, it provides a vehicle for exploring and comparing projection pursuit functions.
This article is from 'Computing Science and Statistics: Proceedings of the Symposium on the Interface Critical Applications of Scientific Computing: Biology, Engineering, Medicine, Speech Held in Seattle, Washington on 21-24 April 1991,' AD-A252 938, p180-183.