Tidy the Result of a Recipe

tidy will return a data frame that contains information regarding a recipe or operation within the recipe (when a tidy method for the operation exists).

# S3 method for recipe
tidy(x, number = NA, id = NA, ...)

# S3 method for step
tidy(x, ...)

# S3 method for check
tidy(x, ...)

Arguments

x	A recipe object (trained or otherwise).
number	An integer or NA. If missing and id is not provided, the return value is a list of the operations in the recipe. If a number is given, a tidy method is executed for that operation in the recipe (if it exists). number must not be provided if id is.
id	A character string or NA. If missing and number is not provided, the return value is a list of the operations in the recipe. If a character string is given, a tidy method is executed for that operation in the recipe (if it exists). id must not be provided if number is.
...	Not currently used.

Value

A tibble with columns that would vary depending on what tidy method is executed. When number and id are NA, a tibble with columns number (the operation iteration), operation (either "step" or "check"), type (the method, e.g. "nzv", "center"), a logical column called trained for whether the operation has been estimated using prep, a logical for skip, and a character column id.

Examples

library(modeldata)
data(okc)

okc_rec <- recipe(~ ., data = okc) %>%
  step_other(all_nominal(), threshold = 0.05, other = "another") %>%
  step_date(date, features = "dow") %>%
  step_center(all_numeric()) %>%
  step_dummy(all_nominal()) %>%
  check_cols(starts_with("date"), age, height)

tidy(okc_rec)
#> # A tibble: 5 x 6
#>   number operation type   trained skip  id          
#>    <int> <chr>     <chr>  <lgl>   <lgl> <chr>       
#> 1      1 step      other  FALSE   FALSE other_OruwB 
#> 2      2 step      date   FALSE   FALSE date_l3nYA  
#> 3      3 step      center FALSE   FALSE center_jp5Kj
#> 4      4 step      dummy  FALSE   FALSE dummy_FhDIk 
#> 5      5 check     cols   FALSE   FALSE cols_IyjzO  

tidy(okc_rec, number = 2)
#> # A tibble: 1 x 4
#>   terms value ordinal id        
#>   <chr> <chr> <lgl>   <chr>     
#> 1 date  dow   FALSE   date_l3nYA
tidy(okc_rec, number = 3)
#> # A tibble: 1 x 3
#>   terms         value id          
#>   <chr>         <dbl> <chr>       
#> 1 all_numeric()    NA center_jp5Kj

okc_rec_trained <- prep(okc_rec, training = okc)
#> Warning: There are new levels in a factor: NA

tidy(okc_rec_trained)
#> # A tibble: 5 x 6
#>   number operation type   trained skip  id          
#>    <int> <chr>     <chr>  <lgl>   <lgl> <chr>       
#> 1      1 step      other  TRUE    FALSE other_OruwB 
#> 2      2 step      date   TRUE    FALSE date_l3nYA  
#> 3      3 step      center TRUE    FALSE center_jp5Kj
#> 4      4 step      dummy  TRUE    FALSE dummy_FhDIk 
#> 5      5 check     cols   TRUE    FALSE cols_IyjzO  
tidy(okc_rec_trained, number = 3)
#> # A tibble: 2 x 3
#>   terms  value id          
#>   <chr>  <dbl> <chr>       
#> 1 age     32.3 center_jp5Kj
#> 2 height  68.3 center_jp5Kj