string sorting in R

sort_polyglot.R

splt <- function(s) {
  x <- strsplit(s, "-")
  s[order(sapply(x, `[[`, 1), as.integer(sapply(x, `[[`, 2)))]
}

splt_radix <- function(s) {
  x <- strsplit(s, "-")
  s[order(sapply(x, `[[`, 1), as.integer(sapply(x, `[[`, 2)), method = "radix")]
}

subs <- function(s) {
  splpos <- regexpr("-", s, fixed = TRUE)
  pt1 <- substr(s, 1, splpos - 1)
  pt2 <- substr(s, splpos+1, 100000)
  o <- order(pt1, as.integer(pt2), method = "radix")
  s[o]
}

set.seed(1)
vec <- expand.grid(
  c("aa", "ab", "ba", "bb", "ac", "ca", "cc"),
  1:200
) |> 
  tidyr::unite(x, sep = "-") |> 
  dplyr::pull() |> 
  sample()

tst <- c("aa-2", "ab-100", "aa-10", "ba-25", "ab-3")

bench::mark(
  splt(tst),
  splt_radix(tst),
  subs(tst)
)
#> # A tibble: 3 × 6
#>   expression           min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>      <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 splt(tst)        23.53µs   25.1µs    38233.        0B     53.6
#> 2 splt_radix(tst)     23µs   24.2µs    40097.        0B     52.2
#> 3 subs(tst)         9.72µs   10.8µs    88700.        0B     53.3

bench::mark(
  splt(vec),
  splt_radix(vec),
  subs(vec)
)
#> # A tibble: 3 × 6
#>   expression           min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>      <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 splt(vec)         1.56ms   1.59ms      623.     131KB    23.7 
#> 2 splt_radix(vec)   1.08ms   1.12ms      876.     131KB    35.5 
#> 3 subs(vec)       165.52µs 173.51µs     5706.      88KB     8.44