Using Color in R
Stowers Institute for Medical Research
R/Bioconductor Discussion Group
Scientific Programmer 17 July 2007
Earl F. Glynn
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
2
Color Basics: Palette
> barplot(rep(1,8), yaxt="n", col=1:8)
Using Color in R
> palette()
[1] "black" "red" "green3" "gray"
3
"blue"
"cyan"
[6] "magenta" "yellow"
Color Basics: Palette
> barplot(rep(1,8), yaxt="n", col=1:8) > barplot(rep(1,8), yaxt="n", col=palette()) > barplot(rep(1,8), yaxt="n", col=c("black", "red", "green3", "blue", "cyan", "magenta", "yellow", "gray"))
Using Color in R
4 Integer color numbers represent offsets into the palette table
Color Basics: Palette
> palette(rainbow(10)) > palette() [1] "red" [6] "cyan" # Redefine palette
Using Color in R
RGB Hex Constants
Color Names
"#FF9900" "#CCFF00" "#33FF00" "#00FF66" "#0066FF" "#3300FF" "#CC00FF" "#FF0099"
> # colors are "recycled" if necessary > barplot(rep(1,20),col=1:20, yaxt="n")
> palette("default") > palette() [1] "black" "blue" "red" "cyan"
Cut/Paste Bitmap
"green3" "gray"
Recycled Colors 5
[6] "magenta" "yellow"
Specify number of colors with rainbow
Color Basics: Palette
> 0:8 / 8 [1] 0.000 0.125 0.250 0.375 0.500 0.625 0.750 0.875 1.000 > palette(gray(0:8 / 8)) > palette() [1] "black" "#202020" "gray25" "#606060" "#808080"
Using Color in R
RGB Hex Constants
[6] "#9F9F9F" "gray75"
"#DFDFDF" "white"
> # colors are "recycled" if necessary > barplot(rep(1,20),col=1:20, yaxt="n") > > palette("default") > palette() [1] "black" "blue" "red" "cyan" "green3" Recycled Colors "gray" 6
[6] "magenta" "yellow"
Specify vector of floats 0.0 to 1.0 with grey/gray
Color Basics: Color Names
> colors() # or colours() [1] "white" [4] "antiquewhite1" [655] "yellow3" > colors()[grep("red", [1] "darkred" [4] "indianred2" [7] "mediumvioletred" [10] "orangered2" [13] "palevioletred" [16] "palevioletred3" [19] "red1" [22] "red4" [25] "violetred2" "aliceblue" "antiquewhite2" "yellow4" colors())] "indianred" "indianred3" "orangered" "orangered3" "palevioletred1" "palevioletred4" "red2" "violetred" "violetred3" "antiquewhite" "antiquewhite3" "yellowgreen"
Using Color in R
"indianred1" "indianred4" "orangered1" "orangered4" "palevioletred2" "red" "red3" "violetred1" "violetred4"
7 Color names from C:\Program Files\R\R-2.5.1\etc\rgb.txt
Color Basics: Color Names
Using Color in R
http://research.stowers-institute.org/efg/R/Color/Chart/index.htm
Color Basics: Color Names
Alphabetical except first color, which is white
Using Color in R
Print seven page table to compare screen colors with printed colors. 9 http://research.stowers-institute.org/efg/R/Color/Chart/ColorChart.pdf
Color Basics: Hex Constants
Index Color Name #rrggbb red green blue
Hex 0 1 2 3 4 5 6 7 8 9 A B C D E F Decimal 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Using Color in R
"Unsaturated" shades of gray
"Saturated" Primary Colors
Hex "FF" = 15*161 + 15*160 = 255 Hex "00" to "FF" can be interpreted as 0.0 to 1.0.
10
Numbers represented in "base 16" are called "hexadecimal". Hex "FF" is largest value represented by one byte (8 bits).
Color Basics
In R Color can be represented by index into palette color name hex constant (24-bit "True Color": 2563 colors = 16,777,216 colors)
Using Color in R
Color Basics
In R many objects can take on different colors: points lines axes text legends background
Using Color in R
Color Basics: Points
> palette()
[1] "black" "red" [6] "magenta" "yellow" "green3" "gray" "blue" "cyan"
4
Using Color in R
> x <- -2:2 > y <- x^2
# Equivalent point colors > plot(x,y, col=1:5, pch=CIRCLE<-16, cex=2)
0 -2
Five (x,y) points to plot
-1
0 x
> plot(x,y, col=c("black", "red", "green3", "blue", "cyan"), pch=CIRCLE, cex=2) > plot(x,y, col=c("#000000", "#FF0000", "green3", 4, 5), pch=CIRCLE, cex=2)
Hex #rrggbb Color name Palette Index
Color Basics: Points
n <- 20 plot(1:n, pch=CIRCLE<-16, cex=1:n, col=rainbow(n))
Using Color in R
1:n
10
15
20
10 Index
15
20
Color Basics: Points
> plot(log2(somite$Signal.1L), log2(somite$Signal.1R))
Using Color in R
p=0
Colors 1 to 32
p=1
> library(gplots) > palette(rev(rich.colors(32)))
How to associate color with p-values at each point?
Color Basics: Points
> palette( rev(rich.colors(32)) ) # colors: 1 to 32 > plot(log2(somite$Signal.1L), log2(somite$Signal.1R), col=1 + 31*somite$worst.p) # col values from 1 to 32
Using Color in R
p=0
p=1
Color Basics: Points and Lines
300
Using Color in R
250
blue
100
150
library(scatterplot3d) ?scatterplot3d # Example 6
200
300 250 200 150 100 50 0 -50 -50 0 50 100 150 200 250 300
-50
red
Red-Green-Blue 3D Plot of colors()
green
50
Color Basics: Axes and Text
BOTTOM LEFT TOP RIGHT <<<<1 2 3 4
100 0 2
Using Color in R
Chart Title 4 6
Top
10 10 Right
par(col.lab="orange", col.main="gray") plot(0:10,0:10, col=0:10, pch=CIRCLE<-16, main="Chart Title", axes=FALSE, xlab="X axis",ylab="Y axis") # Margin Text mtext("Bottom", mtext("Left", mtext("Top", mtext("Right", BOTTOM, LEFT, TOP, RIGHT, col="red") col="green") col="blue") col="magenta") 10 in red
Y axis
80
60 Left
40
axis(BOTTOM, col="red", col.axis="red") # 2, ...
20
AXIS_LABEL_HORIZONTAL <- 1 axis(LEFT, col="green", col.axis="green", at=2*0:5, labels=paste(20*0:5), las=AXIS_LABEL_HORIZONTAL) axis(TOP, col="blue", col.axis="blue") axis(RIGHT, col="magenta", col.axis="magenta")
0 Bottom 0 2 4 X axis 6 8 10
Color Basics: Legends
Time <- 0:120
1.0 Expression -0.5 0.0 0.5
Using Color in R
Period1 <- cos(2*pi*Time/120) Period2 <- cos(2*pi*Time/90) Period3 <- cos(2*pi*Time/150)
COLORS <- c("red", "green", "blue") LINE.TYPES <- c("solid", "dotted", "dashed") Periods <- data.frame( Period1=Period1, Period2=Period2, Period3=Period3)
matplot(Periods, type = "l", xlab="Time[min]",ylab="Expression", col = COLORS, lty = LINE.TYPES) legend("bottomleft", c("120 min period", " 90 min period", "150 min period"), col = COLORS, lty = LINE.TYPES)
-1.0
120 min period 90 min period 150 min period 0 20 40 60 Time[min] 80 100 120
Color Basics: Background
par(bg="light blue") x <- -2:2 y <- x^2 plot(x,y, col=1:5, pch=CIRCLE<-16,cex=2)
y 4
Using Color in R
R's default background color is "transparent."
0 -2
-1
0 x
Set graphic background to simplify cutting and pasting to PowerPoint! Avoids in PowerPoint: Format Picture | Colors and Lines | Fill Color
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
21
Color Space: RGB Color Model
> rgb(1,0,0)
Using Color in R
[1] "#FF0000" > rgb(0,1,0) [1] "#00FF00" > rgb(1,1,0) [1] "#FFFF00" > rgb(255,0,0, maxColorValue=255) [1] "#FF0000" > col2rgb(c("blue", "yellow"))
red green blue [,1] [,2] 0 255 0 255 255 0
#rrggbb
R G B
Color is additive in the RGB Color Model. Coordinate values not always obvious.
22
Red-Green-Blue Color Cube
R = 0 to 255 G = 0 to 255 B = 0 to 255
Using Color in R
Blue
Cyan
255
Magenta
White
255, 255, 255
Not all colors are perceivable on all devices
Grey Line Black
0, 0, 0 255
Red
0.0 to 1.0 0 to 255 (8-bits)
Green 255
Yellow
24-bit graphics: 256 x 256 x 256 = 16,777,216 colors 256 shades of grey
Contrasting Colors?
Color Space: HSV
Match perception of color better than RGB
Using Color in R
Hue-Saturation-Value
Source: http://scien.stanford.edu/class/psych221/projects/02/sojeong/
24
> hsv(1,1,1) [1] "#FF0000" > hsv(1/3,1,1) [1] "#00FF00" > hsv(2/3,1,1) [1] "#0000FF"
Color Space: HSV
Hue-Saturation-Value
HSV (S=1, V=1)
0.275 0.300 0.325 0.350 0.250 0.225 0.200 0.175 0.150 0.125 0.100 0.075 0.050 0.025
Using Color in R
hue <- seq(0.0, 1.0, by=1/40) pie(rep(1,40), labels=formatC(hue, digits=3, format="f"), cex=0.75, col=hsv(hue, 1.0, 1.0), radius=1.0, main="HSV (S=1, V=1)" ) > rgb2hsv(col2rgb("blue")) [,1] h 0.6666667 s 1.0000000 Contrasting Colors? v 1.0000000
0.425 0.450
0.375 0.400
0.475 0.500
0.000 0.975
0.525 0.550 0.575 0.600 0.625 0.650 0.675 0.700 0.725 0.750 0.775 0.800 0.825
0.950 0.925 0.900 0.875 0.850
25
Color Spaces
help(package=colorspace)
Using Color in R
26
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
27
Color Gradients / Color Ramps
par(mfrow=c(6,1), mar=c(3,1,0,1)) BOTTOM <- 1
rainbow
Using Color in R
colorstrip <- function(colors, description, ShowAxis=FALSE) { count <- length(colors) m <- matrix(1:count, count, 1) image(m, col=colors, ylab="", axes=FALSE)
heat.colors
if (ShowAxis) { axis(BOTTOM) } mtext(description, BOTTOM, adj=0.5, line=0.5) }
terrain.colors
topo.colors
COLOR.COUNT <- 256 colorstrip(rainbow(COLOR.COUNT), "rainbow") colorstrip(heat.colors(COLOR.COUNT), "heat.colors") colorstrip(terrain.colors(COLOR.COUNT), "terrain.colors") colorstrip(topo.colors(COLOR.COUNT), "topo.colors") colorstrip(cm.colors(COLOR.COUNT), "cm.colors (cyan-magenta)") colorstrip(gray(0:COLOR.COUNT / COLOR.COUNT),"gray")
cm.colors (cyan-magenta)
gray
28
Color Gradients / Color Ramps
colorRamp and colorRampPalette added in R 2.1.0
1.0 1.0
Using Color in R
m <- outer(1:20,1:20,function(x,y) sin(sqrt(x*y)/3)) rgb.palette <- colorRampPalette(c("red", "orange", "blue"), space = "rgb") filled.contour(m,col = rgb.palette(20))
0.8 0.5
0.6 0.0 0.4
-0.5 0.2
0.0 0.0 0.2 0.4 0.6 0.8 1.0
-1.0
1.0
1.0
# space="Lab" helps when colors don't form a # natural sequence Lab.palette <- colorRampPalette(c("red", "orange", "blue"), space = "Lab") filled.contour(m,col = Lab.palette(20))
0.8 0.5
0.6 0.0 0.4
-0.5 0.2
0.0 0.0 0.2 0.4 0.6 0.8 1.0
29
-1.0
Color Gradients / Color Ramps
HCL (Hue, Chroma, Luminance)
> ?hcl > barplot(rep(1, 20), col = hcl(seq(0, 360, length = 20)))
Using Color in R
See: HCL(Hue-Chroma-Luminance)-based Color Palettes in R http://cran.r-project.org/doc/vignettes/vcd/hcl-colors.pdf
0.0
0.4
0.8
30
Color Gradients / Color Ramps
RColorBrewer Package
> library(RColorBrewer) > display.brewer.all()
YlOrRd YlOrBr YlGnBu YlGn Reds RdPu Purples PuRd PuBuGn PuBu OrRd Oranges Greys Greens GnBu BuPu BuGn Blues Set3 Set2 Set1 Pastel2 Pastel1 Paired Dark2 Accent Spectral RdYlGn RdYlBu RdGy RdBu PuOr PRGn PiYG BrBG
Using Color in R
31
Color Gradients / Color Ramps
Representative Color for Given Wavelength
Using Color in R
Source: http://www.efg2.com/Lab/ScienceAndEngineering/Spectra.htm
32
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
33
Color Blindness
Ishihara Test for Color Blindness
Using Color in R
Normal Color Vision: 25, 29, 45, 56, 6, 8 Red-Green Color Blind: 25, spots, spots, 56, spots, spots
Source: http://www.toledo-bend.com/colorblind/Ishihara.html
About 1 in 12 have some sort of color deficiency: About ~8% of men and ~0.4% of women in the US.
http://en.wikipedia.org/wiki/Color_blindness 34
Color Blindness
R dichromat Package: Color Schemes for dichromats
library(dichromat) par(mfcol=c(1,2)) N <- 20 pie(rep(1,N),col=heat.colors(N)) pie(rep(1,N),col=dichromat(heat.colors(N)))
Using Color in R
dichromat function collapses red-green color distinctions to approximate the effect of the two 35 common forms of red-green color blindness, protanopia and deuteranopia.
Color Blindness
R dichromat Package: Color Schemes for dichromats
library(dichromat) par(mar=c(1,2,1,1)) layout(matrix(1:6,ncol=1)) image(1:10,1,matrix(1:10, ncol=1), col=colorschemes$BrowntoBlue.10, main="Brown to Blue (10)", axes=FALSE) image(1:100,1,matrix(1:100 ,ncol=1), col=colorRampPalette(colorschemes$BrowntoBlue.10,space="Lab")(100), main="Brown to Blue Ramp", axes=FALSE) image(1:10,1,matrix(1:10, ncol=1), col=dichromat(colorschemes$BrowntoBlue.10), main="Brown to Blue (10) -- deuteranopia", axes=FALSE) image(1:12,1,matrix(1:12, ncol=1),col=colorschemes$Categorical.12, main="Categorical (12)", axes=FALSE) image(1:12,1,matrix(1:12, ncol=1), col=dichromat(colorschemes$Categorical.12), main="Categorical (12) -- deuteranopia", axes=FALSE) image(1:12,1,matrix(1:12,ncol=1), col=dichromat(colorschemes$Categorical.12, "protan"), main="Categorical (12) -- protanopia", axes=FALSE)
Using Color in R
36
Color Blindness
R dichromat Package: Color Schemes for dichromats
Brown to Blue (10)
Using Color in R
Brown to Blue Ramp
Brown to Blue (10) -- deuteranopia
Categorical (12)
Categorical (12) -- deuteranopia
Categorical (12) -- protanopia
37
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
38
Why Dont Screen Colors Match Printout?
Using Color in R
Conceptual Diagram of All Colors (2D Slice)
39
Why Dont Screen Colors Match Printout?
Using Color in R
Each Device Has Own Gamut of Colors
40
3D Color Gamut
Using Color in R
From Visualization of Expanded Printing Gamuts Using 3-Dimensional Convex Hulls by Karl Guyler, Hallmark Cards, Kansas City
Color calibration can be used to minimize needless differences
Why Dont Screen Colors Match Printout?
Different
Using Color in R
color gamuts between devices:
How should a color be represented if it doesn't exist on a device?
Imperfect conversions:
CRT screens with RGB (Red-Green-Blue) additive colors must be converted to CMYK (Cyan-Magenta-Yellow-Black) subtractive colors
Color fidelity may not be accurate Paper/ink differences
42
Using Color in R
Color
Basics in R
Color Spaces Color Gradients / Color Ramps Color Blindness
Why Dont Screen Colors Match Printout? Colors Tips
43
Color Tips
Avoid
Using Color in R
unnecessary use of color.
Use bright colors with small graphics to make them stand out. Be consistent in use of color. Don't use color as only attribute to show difference. E.g., consider color and line type.
Also see: Cool Color Commentary, http://www.public-speaking.org/public-speaking-color-article.htm 44
Color Tips: Palettes
Qualitative Palette:
all same perceptual weight/importance typical application: bar plot
Using Color in R
Sequential Palette
for coding numerical information in a range typical application: heat map
Diverging Palette
Like Sequential Palette but with neutral value
Source: Choosing Color Palettes for Statistical Graphics 45 http://epub.wu-wien.ac.at/dyn/virlib/wp/mediate/epub-wu-01_abd.pdf?ID=epub-wu-01_abd
References
Colour for Presentation Graphics
http://www.stat.auckland.ac.nz/~ihaka/colour/color-talk.pdf
Using Color in R
Why should Engineers and Scientists be worried about color?
http://www.research.ibm.com/people/l/lloydt/color/color.HTM
46
