Scaling Numeric Data

Source: R/scale.R

step_scale creates a specification of a recipe step that will normalize numeric data to have a standard deviation of one.

step_scale(
  recipe,
  ...,
  role = NA,
  trained = FALSE,
  sds = NULL,
  factor = 1,
  na_rm = TRUE,
  skip = FALSE,
  id = rand_id("scale")
)

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

Arguments

recipe

A recipe object. The step will be added to the sequence of operations for this recipe.
...

One or more selector functions to choose which variables are affected by the step. See selections() for more details. For the tidy method, these are not currently used.
role

Not used by this step since no new variables are created.
trained

A logical to indicate if the quantities for preprocessing have been estimated.
sds

A named numeric vector of standard deviations. This is NULL until computed by prep.recipe().
factor

A numeric value of either 1 or 2 that scales the numeric inputs by one or two standard deviations. By dividing by two standard deviations, the coefficients attached to continuous predictors can be interpreted the same way as with binary inputs. Defaults to 1. More in reference below.
na_rm

A logical value indicating whether NA values should be removed when computing the standard deviation.
skip

A logical. Should the step be skipped when the recipe is baked by bake.recipe()? While all operations are baked when prep.recipe() is run, some operations may not be able to be conducted on new data (e.g. processing the outcome variable(s)). Care should be taken when using skip = TRUE as it may affect the computations for subsequent operations
id

A character string that is unique to this step to identify it.
x

A step_scale object.

Value

An updated version of recipe with the new step added to the sequence of existing steps (if any). For the tidy method, a tibble with columns terms (the selectors or variables selected) and value (the standard deviations).

Details

Scaling data means that the standard deviation of a variable is divided out of the data. step_scale estimates the variable standard deviations from the data used in the training argument of prep.recipe. bake.recipe then applies the scaling to new data sets using these standard deviations.

References

Gelman, A. (2007) "Scaling regression inputs by dividing by two standard deviations." Unpublished. Source: http://www.stat.columbia.edu/~gelman/research/unpublished/standardizing.pdf.

Examples

library(modeldata)
data(biomass)

biomass_tr <- biomass[biomass$dataset == "Training",]
biomass_te <- biomass[biomass$dataset == "Testing",]

rec <- recipe(HHV ~ carbon + hydrogen + oxygen + nitrogen + sulfur,
              data = biomass_tr)

scaled_trans <- rec %>%
  step_scale(carbon, hydrogen)

scaled_obj <- prep(scaled_trans, training = biomass_tr)

transformed_te <- bake(scaled_obj, biomass_te)

biomass_te[1:10, names(transformed_te)]

#>    carbon hydrogen oxygen nitrogen sulfur    HHV
#> 15  46.35     5.67  47.20     0.30   0.22 18.275
#> 20  43.25     5.50  48.06     2.85   0.34 17.560
#> 26  42.70     5.50  49.10     2.40   0.30 17.173
#> 31  46.40     6.10  37.30     1.80   0.50 18.851
#> 36  48.76     6.32  42.77     0.20   0.00 20.547
#> 41  44.30     5.50  41.70     0.70   0.20 18.467
#> 46  38.94     5.23  54.13     1.19   0.51 15.095
#> 51  42.10     4.66  33.80     0.95   0.20 16.240
#> 55  29.20     4.40  31.10     0.14   4.90 11.147
#> 65  27.80     3.77  23.69     4.63   1.05 10.750
transformed_te

#> # A tibble: 80 x 6
#>    carbon hydrogen oxygen nitrogen sulfur   HHV
#>     <dbl>    <dbl>  <dbl>    <dbl>  <dbl> <dbl>
#>  1   4.45     4.74   47.2     0.3    0.22  18.3
#>  2   4.16     4.60   48.1     2.85   0.34  17.6
#>  3   4.10     4.60   49.1     2.4    0.3   17.2
#>  4   4.46     5.10   37.3     1.8    0.5   18.9
#>  5   4.68     5.28   42.8     0.2    0     20.5
#>  6   4.26     4.60   41.7     0.7    0.2   18.5
#>  7   3.74     4.37   54.1     1.19   0.51  15.1
#>  8   4.04     3.89   33.8     0.95   0.2   16.2
#>  9   2.81     3.68   31.1     0.14   4.9   11.1
#> 10   2.67     3.15   23.7     4.63   1.05  10.8
#> # … with 70 more rows
tidy(scaled_trans, number = 1)

#> # A tibble: 2 x 3
#>   terms    value id         
#>   <chr>    <dbl> <chr>      
#> 1 carbon      NA scale_uY1pe
#> 2 hydrogen    NA scale_uY1pe
tidy(scaled_obj, number = 1)

#> # A tibble: 2 x 3
#>   terms    value id         
#>   <chr>    <dbl> <chr>      
#> 1 carbon   10.4  scale_uY1pe
#> 2 hydrogen  1.20 scale_uY1pe