(Builds on: The grammar of graphics)
Summary functions take a vector of data and return a single number; you
use them in conjunction with dplyr::summarise(). The ggplot2 equivalent
are the summary geoms, which split the data into pieces, apply some summary
and visualise the results.
Summaries are useful when you have a lot of data (and plotting it all leads to overplotting), or when the data is noisy (and you need to make small but consistent patterns more clear).