Augments dplyr::top_n() to also return a summary of the remainder of the dataset that didn't make the cut. Useful when working with proportions.

example.md

library(magrittr)
library(dplyr)
library(stringr)

df <- mtcars; df$car <- rownames(df)

df %>%
  mutate(make = str_extract(car, "^[A-Za-z]+\\b")) %>%
  group_by(make) %>%
  summarize(n = n()) %>%
  mutate(proportion = n/sum(n)) %>%
  arrange(desc(proportion)) %>%
  top_n2(5, proportion, make) %>%
  select(-n) %>%
  mutate(proportion = sprintf("%.2f%%", 100*proportion)) %>%
  knitr::kable()

make	proportion
Merc	21.88%
Fiat	6.25%
Hornet	6.25%
Mazda	6.25%
Toyota	6.25%
...17 others...	53.12%

top_n2.R

library(magrittr)
library(dplyr)

top_n2 <- function(x, n, wt, var, f = sum) {
  expr <- paste0("dplyr::top_n(x, ", n, ", ", deparse(substitute(wt)), ")")
  y <- eval(parse(text = expr))
  `%notin%` <- function(x, y) {
    return(!(x %in% y))
  }
  filter_dots <- eval(parse(text = paste0("list(~", deparse(substitute(var)), "%notin% y$", deparse(substitute(var)), ")")))
  summary_dots <- eval(parse(text = paste0("list(~paste0('...', length(", deparse(substitute(var)),
                                           "), ' others...'), ~f(", deparse(substitute(wt)), "))")))
  z <- x %>%
    filter_(.dots = filter_dots) %>%
    summarize_(.dots = summary_dots)
  names(z) <- c(deparse(substitute(var)), deparse(substitute(wt)))
  return(bind_rows(y, z))
}