default.colors: Color schemes

In paulemms/datamining: Data mining

Color schemes

Description

Compute a color scheme with a specified number of levels.

Usage

default.colors(n)

Arguments

n	the number of colors desired

Details

The first two schemes are categorical, providing maximum separation in hue, intended for depicting unordered categories. default.colors has only dark colors, good for coloring points, while default.colors.w includes light colors, good for filling regions.

The next four schemes are sequential, from light to dark, with variation in hue to increase discrimination. They are intended for depicting ordered levels. The sequential order is more easily perceived with these schemes than with the built-in palettes heat.colors, terrain.colors, and topo.colors. The ordering can also be seen by the color-blind and when printed in black and white.

The main difference between the sequential schemes is the variation in hue, with YR.colors having the most variation and gray.colors having the least. Generally you should choose the amount of variation according to the number of levels. I recommend OrRd.colors for three levels, YlGnBu.colors for four to eight levels, and YR.colors beyond eight levels.

The last four schemes are double-ended or diverging schemes, which progress from one hue to a second hue, passing through white in the middle. They are intended for representing signed ordered levels, such as residuals. The main difference between them is the amount of separation between colors, so generally you use GM.colors when you want a few levels and RYB.colors when you want many.

These functions can be used as the color.palette parameter to filled.contour and color.plot, for example.

Value

A vector of strings, naming colors.

Author(s)

Tom Minka

References

The schemes in YlGnBu.colors, OrRd.colors, RYB.colors, and BrBg.colors are from ColorBrewer. The scheme in YR.colors is from Howard Seltman.

Mark A. Harrower and Cynthia A. Brewer. ColorBrewer: An Online Tool for Selecting Color Schemes for Maps, The Cartographic Journal, in press. http://www.colorbrewer.org/, http://www.personal.psu.edu/faculty/c/a/cab38/ColorBrewerBeta.html

Generalized color schemes for Mapping and Visualization. From Cynthia Brewer, Color Use Guidelines for Mapping and Visualization. Reprinted at the Gallery of Data Visualization by Michael Friendly. http://www.math.yorku.ca/SCS/Gallery/images/S12-fullstructureClean.gif

Dan Carr. Color perception, the importance of gray and residuals, on a choropleth map. Statistical Computing & Graphics Newsletter 5(1):16-20, 1994 http://cm.bell-labs.com/cm/ms/who/cocteau/newsletter/issues/back/v51.pdf

See Also

colors,rainbow,color.cone

Examples

data(Housing)
color.plot(Price ~ Rooms + Low.Status, Housing, bg=gray(0.5),
          color.palette=YlGnBu.colors)
color.plot(Price ~ Rooms + Low.Status, Housing, bg=gray(0.5),
            color.palette=YR.colors)
color.plot(Price ~ Rooms + Low.Status, Housing, bg=gray(0.5),
            color.palette=RYB.colors,nlevels=5)

# also see examples for color.cone

paulemms/datamining documentation built on March 1, 2023, 4:01 p.m.