Search code examples
rdplyrrlang

How can I use NSE and filter an expression using ... (ellipsis)?

How can I utilize the magic dots (...) / ellipsis in order to filter based off an arbitrary column?

df = tibble::tibble(col1 = c('a', 'b', 'c'), col2 = c(1,3,4))

my_func = function(x, ...){
  df %>%
    dplyr::filter(... == x)
}

my_func('a', col1)


# Should return:
# A tibble: 1 x 2
  col1   col2
  <chr> <dbl>
1 a         1
Solution

  • We can convert to a quosure (quo) and evaluate (!!). Here, we assume there would be only a single column passed into the filter

    my_func <- function(x, ...){
   df %>%
      dplyr::filter(!! quo(...) == x)
  }



my_func('a', col1)
# A tibble: 1 x 2
#  col1   col2
#  <chr> <dbl>
#1 a         1

    If there are multiple columns, then it may be better to use filter_at

    It may be also better to make use of {{}} if we can pass as an argument instead of ...

    my_func <- function(x, coln) {
        df %>%
              filter({{coln}} == x)
    }
my_func('a', col1)