performance comparison of table functions in R

table-performance.R

See https://twitter.com/malte_grosser/status/1262862749794795520?s=20

``` r
table2 <- function(x) {
  x_u <- if (is.factor(x)) sort(x[!duplicated(x)]) else sort(unique(x))
  x_f <- factor(x, levels = as.character((x_u)), exclude = NULL)
  t_x_f <- tabulate(x_f)
  names(t_x_f) <- as.character(x_u)
  t_x_f
}

table3 <- function(x) {
  x_u <- if (is.factor(x)) x[!duplicated(x)] else unique(x)
  x_f <- factor(x, levels = as.character((x_u)), exclude = NULL)
  t_x_f <- tabulate(x_f)
  data.frame(values = x_u, count = t_x_f[seq_along(x_u)])
}

set.seed(123)
x <- paste0(letters, sample(500, size = 25000, replace = TRUE))

bench::mark(
  table(x),
  table2(x),
  table3(x),
  check = FALSE,
  iterations = 100
)[1:7]
#> # A tibble: 3 x 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 table(x)     80.5ms     90ms      10.6    2.93MB    0.560
#> 2 table2(x)    78.3ms   86.8ms      11.2    1.59MB    0.345
#> 3 table3(x)    80.9ms   94.6ms      10.2    2.96MB    0.652


x <- sample(1:100, size = 10000, replace = T)
set.seed(123)
bench::mark(
  table(x),
  table2(x),
  table3(x),
  check = FALSE,
  iterations = 100
)[1:7]
#> # A tibble: 3 x 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 table(x)      680us    903us     1037.     689KB     10.5
#> 2 table2(x)     488us    496us     1862.     409KB     18.8
#> 3 table3(x)     652us    663us     1474.     409KB      0
```

<sup>Created on 2020-05-22 by the [reprex package](https://reprex.tidyverse.org) (v0.3.0)</sup>