dplyr:: How to use SE and `%in%` in `filter`?

Question

Consider

df = data.frame(col1 = 1:3, col2 = 2:4)

I want to do

df %>%
  filter(col1 %in% 2:3)

which should return the 2nd and 3rd row, but col1 is passed in as a string as variable str_1 = "col1"

So I tried

  str_1 = "col1"
  df %>% 
    filter(!!!str_1 %in% 2:3)

  df %>% 
    filter(!!str_1 %in% 2:3)

and niether worked. What's the right syntax for this? And what's an easier way to remember these dplyr tricks?

`df %>% filter(!!sym(str_1) %in% 2:3)` – hmhensen Dec 21 '19 at 02:24 — hmhensen, Dec 21 '19 at 02:24

Ronak Shah · Answer 1 · 2019-12-21T02:34:02.577

1

You can use filter_at when the you have column names as strings.

library(dplyr)
df %>% filter_at(str_1, any_vars(. %in% 2:3))
#This seems to work as well
#df %>% filter_at(str_1, ~. %in% 2:3)

#  col1 col2
#1    2    3
#2    3    4

filter_at like all other _at variants in dplyr has a .vars arguments which accepts a character vector of column names or positions to apply a function.

edited Dec 21 '19 at 02:34

answered Dec 21 '19 at 02:27

Ronak Shah

377,200
20
156
213

explain what filter_at means? – xiaodai Dec 21 '19 at 02:29
@xiaodai Added some explanation. – Ronak Shah Dec 21 '19 at 02:34

dplyr:: How to use SE and `%in%` in `filter`?

1 Answers1