A somewhat contrived example . . . first, illustrate the problem:
library(ggplot2)
ggplot(iris)+aes(x=Sepal.Width)+geom_histogram()+facet_grid(Species~.)
How do we get these histograms to be better sorted? The following will reorder the factor variable, Species, by the mean of Sepal.Width:
iris$Species = reorder(iris$Species, iris$Sepal.Width, mean)
ggplot(iris)+aes(x=Sepal.Width)+geom_histogram()+facet_grid(Species~.)
Now the histograms are sorted by the mean sepal width.
I hope that in a future version you’ll be able to do
ggplot(iris)+aes(x=Sepal.Width)+geom_histogram()+facet_grid(reorder(Species, Sepal.Width) ~.)
directly.