Methods for the graphic representation of systems simulated data

Abstract

This paper explores the use of three multidimensional graphic techniques as a means of quick and easy exploration of computer-simulated data. Harmonic function plots, facial plots and three-dimensional box plots are presented as tools for quickly assessing simulated data structures: These techniques are then experimentally compared to assess their efficacy in recovering simulated data clusters. Findings indicate that the techniques vary in their ability to correctly recover clusters. The factor of user familiarity with the type of graphic representation used seems to be most influential in obtaining true insight into the data.