R recipes step_regex 检测正则表达式

step_regex(
  recipe,
  ...,
  role = "predictor",
  trained = FALSE,
  pattern = ".",
  options = list(),
  result = make.names(pattern),
  input = NULL,
  keep_original_cols = TRUE,
  skip = FALSE,
  id = rand_id("regex")
)

其他虚拟变量和编码步骤：step_bin2factor() , step_count() , step_date() , step_dummy_extract() , step_dummy_multi_choice() , step_dummy() , step_factor2string() , step_holiday() , step_indicate_na() , step_integer() , step_novel() , step_num2factor() , step_ordinalscore() , step_other() , step_relevel() , step_string2factor() , step_time() , step_unknown() , step_unorder()

data(covers, package = "modeldata")

rec <- recipe(~description, covers) %>%
  step_regex(description, pattern = "(rock|stony)", result = "rocks") %>%
  step_regex(description, pattern = "ratake families")

rec2 <- prep(rec, training = covers)
rec2
#> 
#> ── Recipe ────────────────────────────────────────────────────────────────
#> 
#> ── Inputs 
#> Number of variables by role
#> predictor: 1
#> 
#> ── Training information 
#> Training data contained 40 data points and no incomplete rows.
#> 
#> ── Operations 
#> • Regular expression dummy variable using: "(rock|stony)" | Trained
#> • Regular expression dummy variable using: "ratake families" | Trained

with_dummies <- bake(rec2, new_data = covers)
with_dummies
#> # A tibble: 40 × 3
#>    description                                       rocks ratake.families
#>    <fct>                                             <int>           <int>
#>  1 1,cathedral family,rock outcrop complex,extremel…     1               0
#>  2 2,vanet,ratake families complex,very stony            1               1
#>  3 3,haploborolis,rock outcrop complex,rubbly            1               0
#>  4 4,ratake family,rock outcrop complex,rubbly           1               0
#>  5 5,vanet family,rock outcrop complex complex,rubb…     1               0
#>  6 6,vanet,wetmore families,rock outcrop complex,st…     1               0
#>  7 7,gothic family                                       0               0
#>  8 8,supervisor,limber families complex                  0               0
#>  9 9,troutville family,very stony                        1               0
#> 10 10,bullwark,catamount families,rock outcrop comp…     1               0
#> # ℹ 30 more rows
tidy(rec, number = 1)
#> # A tibble: 1 × 3
#>   terms       result id         
#>   <chr>       <chr>  <chr>      
#> 1 description NA     regex_0cUH6
tidy(rec2, number = 1)
#> # A tibble: 1 × 3
#>   terms       result id         
#>   <chr>       <chr>  <chr>      
#> 1 description rocks  regex_0cUH6