Online Program Home
My Program

Abstract Details

Activity Number: 392
Type: Roundtables
Date/Time: Tuesday, August 2, 2016 : 12:30 PM to 1:50 PM
Sponsor: Section on Statistical Graphics
Abstract #320554
Title: The Use of Color in Statistical Graphics
Author(s): Kevin Keen*
Companies: University of Northern British Columbia
Keywords: color ; color perception ; color space ; graph drawing ; graph coloring ; graphicacy
Abstract:

Selecting optimal colors for a graphical display is nontrivial and depends on the choice of media. The default color space in the R software system is sRGB. Color printers use either the CMYK or CcMmYK color models. Software and hardware exists to harmonize LCD and printed colors. Humans can identify up to 8 million colors, but studies in statistical graphics have shown that color is a poor choice for representing a continuous variable. Humans do not order color along a linear scale. While color scales can be learned, the HCL and CIELUV color spaces are better alternatives to the rainbow and heat-map models. On the other hand, grayscale provides a consistent and effective ordering that is easily reproducible in print. Functions in the grDevices package in R can produce sRGB, CMYK, and grayscale output. About 1 in 12 men have some form of color-viewing deficiency. The dichromat package in R can be used to approximate the effects of dichromacy. The website colorbrewer2.org permits checking a range of color pallets as whether a given pallet is colorblind-, LCD-, or printer-friendly. In any event, it is best to duplicate information carried by color with different symbols or ordering.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association