Visual Analytics to Identify Temporal Patterns and Variability in Simulations from Cellular Automata

Abstract

Cellular Automata (CA) are discrete simulation models, thus producing spatio-temporal data through experiments, as well as stochastic models, thus generating multi-run data. Identifying temporal patterns, such as cycles, is important to understand the behavior of the model. Assessing variability is also essential to estimate which parameter values may require more runs and what consensus emerges across simulation runs. However, these two tasks are currently arduous as the commonly employed slider-based visualizations offer little support to identify temporal trends or excessive model variability. In this article, we addressed these two tasks by developing, implementing, and evaluating a new visual analytics environment that uses several linked visualizations. Our empirical evaluation of the proposed environment assessed (i) whether modelers could identify temporal patterns and variability, (ii) how features of simulations impacted performances, and (iii) whether modelers can use the familiar slider-based visualization together with our new environment. Results shows that participants were confident on results obtained using our new environment. They were also able to accomplish the two target tasks without taking longer than they would with current solutions. Our qualitative analysis found that some participants saw value in switching between our proposed visualization and the commonly used slider-based version. In addition, we noted that errors were affected not only by the type of visualizations but also by specific features of the simulations. Future work may combine and adapt these visualizations depending on salient simulation parameters.