What's the best way to find the minimum over all columns of strings in a tibble?

user_16911683 0
Viewed 59

Suppose I have a tibble such as

t <- tibble(
  x = c("a", "b", "c"),
  y = c("d", "e", "f"),
  z = c("g", "h", "i")
)

and I would like to find the minimum value over all values in all columns. What's the best way to do this within the Tidyverse mindset?

If I try min(t), I get the error

Error in FUN(X[[i]], ...) : 
  only defined on a data frame with all numeric-alike variables

The only ways I've found to do this so far involve extracting out the individual vectors and finding the min over those. For example, t_min <- min(c(t$x, t$y, t$z)) gives "a". This works, but it doesn't extend well to many columns.

Note: assume that the "min" function gives the desired result when applied to the string data.

tibble r
4 Answers

With tidyverse/dplyr you can

library(tidyverse)

t %>% 
  summarise(across(everything(), min))

# A tibble: 1 × 3
  x     y     z    
  <chr> <chr> <chr>
1 a     d     g

Minimum of the dataset

t %>% 
  summarise(across(everything(), min)) %>% 
  rowwise() %>% 
  mutate(min = min(c_across(x:z)))
user_172133550

Here's how to get the min value from each column (thanks @benson for the clean implementation):

sapply(t, min)
#   x   y   z 
# "a" "d" "g"

And how to get the min value from an entire tibble (thanks to @m-- for pointing out that it's not necessary to unname() the vector before taking the min()):

min(unlist(t))
# [1] "a"
user_57837450

Another possible solution, based on apply:

library(tibble)

apply(t, 2, min)

#>   x   y   z 
#> "a" "d" "g"
user_115644870

You can use rlang::exec() to splice your columns into a call to c(), and then call min() on the result:

min(rlang::exec(c, !!!t))
#> [1] "a"

A similar approach in base R is to use do.call():

min(do.call(c, t))
#> [1] "a"
user_196510150
Related