Question

Is it possible to filter out subsets of the data that have small numbers of observations within a ggplot2 call?

For example, take the following plot: qplot(price,data=diamonds,geom="density",colour=cut)

The plot is a little busy, and I would like the exclude the cut values with a small number of observations, ie,

> xtabs(~cut,diamonds)
cut
     Fair      Good Very Good   Premium     Ideal 
     1610      4906     12082     13791     21551

the Fair and Good qualities of the cut factor.

I’m wanting a solution that can fit an arbitrary data set and if possible be able to select not just by a threshold number of observations, but by top 3 for example.

Answer 1

ggplot(subset(diamonds, cut %in% arrange(count(diamonds, .(cut)), desc(freq))[1:3,]$cut),
  aes(price, colour=cut)) + 
  geom_density() + facet_grid(~cut)

1. count counts up each elements into data.frame.
2. arrange orders a data.frame based on the specified column.
3. desc enables reversed-order sorting.
4. finally subset the rows whose cut is included in the top 3 by %in%.