Interactive Visualization of Trees Using R and D3

• Activity Number: 158
• Type: Invited
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Graphics
• Author(s): Kenneth Shirley*
• Companies: AT&T Labs Research
• Keywords: Trees ; Interactive Graphics ; D3 ; Dimension Reduction

Abstract

We discuss methods for interactively visualizing trees, and, in particular, smaller versions of those trees known as "summary trees" which enable exploratory data analysis on large hierarchical data sets. We begin by introducing "summary trees", which are small trees that summarize the structure of a larger tree by aggregating its nodes according to certain constraints, and then optimizing a certain measure of the resulting aggregated tree. Next, we demonstrate how to interactively visualize such trees using R and D3. Our example data sets will include trees representing internet traffic data, phylogenetic trees, organizational charts, and others. We will alternate between answering specific questions about data using graphical tools, and discussing general principles of interactive graphics in R and D3.