ggplot(iris[iris$Species == "setosa",],aes(Sepal.Width)) + geom_density()
Here, we are subsetting the dataframe before passing it to ggplot. It is a very useful tool derived from the data frame data structure.
To make the code more readable, one can also use
library(dplyr) iris %>% filter(Species == "setosa") %>% ggplot(aes(Sepal.Width)) + geom_density()