Count occurence of values for every possible pair

Question

I have a list of ids and places where these ids have been. Now I want to find pairs of ids that have most places in common.

My data frame looks like this:

Answer 1

For a dplyr-esque solution,

You could do:

left_join(df, df, by = "place") %>%
  rename(pair1 = id.x, pair2 = id.y) %>%
  filter(!pair1 == pair2, !duplicated(t(apply(., 1, sort))) == TRUE) %>% 
  count(pair1, pair2)