Visualizing Statistical Properties of Smoothly Brushed Data Subsets

In many application fields, the statistical properties of data sets are of great interest for data analysts. Since local variations can occur especially in large datasets, it is useful to visualize not only global values, but also the properties of user-defined subsets. Hence, we present in this paper a visualization of the statistical characteristics of subsets, with an emphasis on the temporal development. The visualization is coordinated with other views. They provide details about the data and allow for smooth brushing of subsets, a concept that introduces a continuous transition from data in focus to context data. Our approach accounts for smooth brushing in both the derivation and the visualization of statistical properties in order to visualize variations within subsets. To analyze differences between multiple subsets, our approach further integrates visualization concepts for the comparison of statistical properties. An application example from the simulation of biological systems demonstrates the benefits of our approach.