Assume the following function
library(tidyverse)
nest_fun <- function(data, groups) {
data %>%
group_by(pick({{groups}})) %>%
nest() %>%
mutate(filename = paste({{groups}}[1], {{groups}}[2], "file.png", sep = "_"))
}
nest_fun(data = mtcars, groups = c(vs, am))
# # A tibble: 4 × 4
# # Groups: vs, am [4]
# vs am data filename
# <dbl> <dbl> <list> <chr>
# 1 0 1 <tibble [6 × 9]> 0_1_file.png
# 2 1 1 <tibble [7 × 9]> 1_1_file.png
# 3 1 0 <tibble [7 × 9]> 1_0_file.png
# 4 0 0 <tibble [12 × 9]> 0_0_file.png
The function works fine as long as I group by exactly 2 variables. How could I write the function such that I can group by an arbitrary number of variables and still get filename correct?
The collapse argument of paste (and paste0) pastes down the elements of a vector, whereas sep pastes across the arguments of the call. Therefore you can get what you want with nested calls to paste/paste0.
I strongly support @r2evans' comment about providing the data frame as an argument to your function. It is simply good practice (and allows you to pipe your function if you wish). I see you've already actioned that in the question. However, using data - which is the name of a built-in function - as its name is not ideal.
So...
library(tidyverse)
nest_fun <- function(d=mtcars, groups = c(vs, am)) {
d %>%
group_by(pick({{groups}})) %>%
nest() %>%
mutate(filename = paste0(paste({{groups}}, collapse="_"), "_file.png"))
}
nest_fun()
# A tibble: 4 × 4
# Groups: vs, am [4]
vs am data filename
<dbl> <dbl> <list> <chr>
1 0 1 <tibble [6 × 9]> 0_1_file.png
2 1 1 <tibble [7 × 9]> 1_1_file.png
3 1 0 <tibble [7 × 9]> 1_0_file.png
4 0 0 <tibble [12 × 9]> 0_0_file.png
mtcars %>% nest_fun(c(vs, am, gear))
# A tibble: 7 × 5
# Groups: vs, am, gear [7]
vs am gear data filename
<dbl> <dbl> <dbl> <list> <chr>
1 0 1 4 <tibble [2 × 8]> 0_1_4_file.png
2 1 1 4 <tibble [6 × 8]> 1_1_4_file.png
3 1 0 3 <tibble [3 × 8]> 1_0_3_file.png
4 0 0 3 <tibble [12 × 8]> 0_0_3_file.png
5 1 0 4 <tibble [4 × 8]> 1_0_4_file.png
6 0 1 5 <tibble [4 × 8]> 0_1_5_file.png
7 1 1 5 <tibble [1 × 8]> 1_1_5_file.png
nest_fun()