CODEWITHSUNDEEP
rplotlattice
dewarrn1
dewarrn1

Reputation: 391

How do I vary colors on a non-grouping factor in R's lattice barcharts?

I'm trying to plot data in a barchart using lattice graphics. I'd like to sort the bars by one factor and group by another (i.e., position by the first factor and contrast within position by the second). However, I'd like to color the bars by the first (i.e., position) factor.

In the following example, the plot is rendered with paired red and blue bars. Instead, I'd like two adjacent red bars, two adjacent blue bars, and two adjacent green bars.

library(lattice)

data = runif( 6 )
facA = rep( c( "a", "b", "c" ), each = 2 )
facB = rep( c( "1", "2" ), 3 )
df = data.frame( "FactorA" = facA, "FactorB" = facB, "Data" = data )
df$FactorA = = as.factor( FactorA )
df$FactorB = = as.factor( FactorB )
test_colors = c( "red", "blue", "green" )
test_plot = barchart(FactorA ~ Data, groups = FactorB, data = df, 
                     horizontal = TRUE, col = test_colors, origin = 0 )
plot( test_plot )

Thanks in advance!

Upvotes: 2

Views: 749

Answers (2)

Dinre
Dinre

Reputation: 4216

If you wanted to do this with 'ggplot2', you would write it like so:

ggplot(df, aes(x=FactorA, y=data, group=FactorB, fill=FactorA)) +
geom_bar(stat="identity", position="dodge") + 
coord_flip()

From there, it's just cosmetic tailoring to suit your tastes, including picking the colors for the fill.

Upvotes: 0

plannapus
plannapus

Reputation: 18759

This is not perfect but I think it does the trick (if I understood correctly your question).

barchart( FactorB ~ Data | FactorA, data = df, groups=FactorA, 
          horizontal = TRUE, col = test_colors, origin = 0,  layout=c(1,3))

Factor B ~ Data | FactorA means it divides your data in panels corresponding to FactorA and inside each of these group, split according to FactorB. The color follows what have been defined as the groups.

enter image description here

Upvotes: 2

Related Questions