how to filter out rarities in R

filter_rarities.R

library(dplyr)

d <- data_frame(A =c(rep("a", 5), rep("c",2)))
d

## one way to do it

d %>% 
  group_by(A) %>% 
  filter(length(A) > 2)

## longer, but easier on the eyes.

rare_groups <- d %>% 
  group_by(A) %>% 
  tally %>% 
  filter(n <= 2)

## examine rare_groups to observe which groups are called "rare"
## remove them from original data with an "anti join":

d %>% 
  anti_join(rare_groups)