Using dplyr to sort a table by two alphanumeric columns

Question

I have a dataset that looks like this:

library(dplyr)  

Data <- tibble(
      Area1 = rep(c("A1 1AA", "B3 4TT","D1 1AA", "A10 6TY","A2 9GG"),2),
      Area2 = c("A2 7BB", "B11 5TT","A14 9SS","A4 4HH","V6 9FF", "A11 6TT","B4 3DD","D1 4FF","G5 7DD","A2 7YY"))

I would like to sort it by Area1 and then Area2 , however arrange does not produce the desired result because it's in lexicographical order.

Data %>% arrange(Area1,Area2) #not the desired order

Is there a way using dplyr to produce this output that is in the desired order?

Output <- tibble(
      Area1 = c("A1 1AA", "A1 1AA", "A2 9GG","A2 9GG","A10 6TY","A10 6TY", "B3 4TT","B3 4TT","D1 1AA","D1 1AA"),
      Area2 = c("A2 7BB", "A11 6TT","A2 7YY","V6 9FF","A4 4HH", "G5 7DD","B4 3DD","B11 5TT","A14 9SS","D1 4FF"))

Answer 1

Seems like we can use mixedorder with slice

library(dplyr)
library(gtools)
library(stringr)
Output2 <- Data %>% 
              slice(mixedorder(str_c(Area1, Area2)))

---

Or another option is to remove the numeric, non-numeric separately and use that in arrange

Output3 <- Data %>%
    arrange(str_remove_all(Area1, "\\d+"), 
            readr::parse_number(Area1),
            str_remove_all(Area2, "\\d+"), 
           readr::parse_number(Area2)) 

-checking with OP's expected

identical(Output, Output2)
#[1] TRUE

identical(Output, Output3)
#[1] TRUE

Answer 2

Here is another option using arrange() and str_sort() :

library(dplyr)
library(stringr)

Data %>%
  arrange(across(starts_with("Area"), ~match(.x, str_sort(unique(.x), numeric = TRUE))))

# A tibble: 10 x 2
   Area1   Area2  
   <chr>   <chr>  
 1 A1 1AA  A2 7BB 
 2 A1 1AA  A11 6TT
 3 A2 9GG  A2 7YY 
 4 A2 9GG  V6 9FF 
 5 A10 6TY A4 4HH 
 6 A10 6TY G5 7DD 
 7 B3 4TT  B4 3DD 
 8 B3 4TT  B11 5TT
 9 D1 1AA  A14 9SS
10 D1 1AA  D1 4FF