Personnel

Class meetings

Hamilton Hall 207: - Mon 10:30-11:20 - Tue 1:30-3:20

General info

Assignments and assessment

Course objective and content

The objective of this course is for students to gain both a theoretical and practical grounding in the issues surrounding data visualization in statistics and data science, including both the foundational work of Cleveland, Tufte, and Wilkinson and recent developments and controversies in the area of data visualization. For practical applications the course will focus on the ggplot package for R, but other more specialized platforms (e.g. ggobi, leaflet, D3.js) will also be discussed.

The first half of the course will consist of lectures and demonstrations presented by the instructors. The second half of the course will shift to a student-led model, where students give presentations focusing on various advanced topics and work on a group project.

part 1: core topics

  • data manipulation basics for data visualization
  • graphical principles
    • Cleveland: perception and hierarchy
    • Tufte: minimalism in data presentation
    • Wilkinson and Wickham: the grammar of graphics
  • graphics for quality assurance and exploratory data analysis
  • graphics for model diagnostics
  • graphics for inference (coefficient plots etc.)
  • ethical issues in data visualization
  • approaches to big data
  • colour theory

part 2: specialized/student-led topics

(which of these topics are covered and in what order will depend on student feedback)

  • info viz vs. data viz: chartjunk and graphics purists
  • perspective plotting (rgl, plotly)
  • high-dimensional data (ggobi, ggvis etc.)
  • dynamic graphics (ggvis, plotly, google charts, D3.js (?))
  • spatial data and mapping
  • smoothing for graphical display
  • alternative platforms