Edward Tufte
Values minimalism, showing relationships with as little ink as possible
Emphasizes the story behind the data, vizualizers should show causality, similarities, and differences
Principle of small multiples: show multiple smaller graphs for comparison rather than plotting on a single plot
Prefers high-density graphs to show as much information in as little space as possible
Don’t “lie” to your audience by presenting proportions in visualization that are more exaggerated than the data itself
Hadley Wickham
“Tidy data” where rows are observations and columns are different variables
Visualizations should take information from data to more easily interpreted output
Code should be used to make visualizations, especially because it is easy to debug as you can easily see every step taken
Reproducability of visualizations is very important, and is aided by using code
Create simple visualizations to find the right type, then create a higher effort plot of that type
Comparison
Tufte’s primary focusis on cleaning out “chart junk” where Wickham’s is getting pen-and-paper visualizers to use code for their work
Both Tufte and Wickham stress the need for visualizations to be correctly interpreted by the viewer
Wickham’s principle of creating simple plots during the discovery process aligns with Tufte’s need for minimalist plots, though Wickham stressed that more complex pots could be created later in the process
Tufte’s and Wickhams design principles can easily work together, as “tidy data” can be implemented with code to create small multiples of simplistic minimal-“ink” plots that are both easily and accurately interpreted by the viewer