STA 313 - Advanced Data Visualization is all about the art and science of visualizing data. Three themes (what, why, and how) will run alongside each other as we cycle through the course topics. In “what” we focus on specific types of visualizations for a particular purpose (e.g. maps for spatial data, Sankey diagrams for proportions, etc.) as well as the tooling to produce them (e.g. specific R packages). In “how” we focus on the process – each visualization starts with a design (which we’ll often ask you to do with a rough sketch accompanied by pseudo code), then often needs pre-processing of the data (wrangling, reshaping, joining, etc. to get it into a tidy, rectangular format for visualization), then attributes of the data are mapped to plot aesthetics, then the creator of the visualization needs to make a series of strategic decisions about visual encoding (e.g. accessibility concerns), and finally creating effective visualizations requires post-processing for visual appeal as well as annotation. In “why” we discuss the theory that ties the “how” and the “what” together, often focusing on the grammar of graphics. Like any data analysis, data visualization is also an iterative process. We don’t expect you to land on the perfect visualization on the first try, so we promote the iterative process via critical and constructive review of one’s own and each others’ work. Independent modules will also touch on topics such as using statistical graphics for visual inference, creating data-based art, and a review of the literature on non-visual approaches to representing data.
The course will focus on the use of the R statistical programming language and introduce you to a variety of modern data visualization packages in R. In addition, you will continue to use hone their data science workflow skills that they acquired in pre-requisite courses by working with Git and GitHub for version control and collaboration.