ABSTRACT
This chapter explains a basic strategy for the analysis step and one or two selected techniques that implement the strategy. It discusses the approach that addresses the problem of visual clutter, this time, by bundling geometrical primitives. The chapter focuses on decluttering such visual representations by summarizing geometrical line primitives into so-called bundles. It discusses the basic steps of feature-based visual analysis. The chapter also discusses the feature-based approach further in the context of a particularly challenging problem: the analysis of chaotic movement. It provides an overview of computational approaches that can support the analysis of large and complex data. The overview supports the exploration of movement features and parameter dependencies across all simulation runs. The chapter argues that automatic computations, principal component analysis (PCA) in this case, can support the interactive visual data analysis. PCA is a helpful approach to determine which data variables play an important role and which can potentially be neglected.
