Skip to content

Pre-processing Analysis Data

create_recipe.Rd

Relative too the recipes package, the recipes::recipe() and recipes::prep() functions are combined.

Usage

create_recipe(
  data,
  feat = NULL,
  log10 = TRUE,
  center = TRUE,
  scale = TRUE,
  ...
)

bake_recipe(x, data)

is.baked(data)

convert_recipe(object)

Arguments

data

A data.frame to be pre-processed according to x.

feat

character(n). A vector indicating which variables to center/scale.

log10

logical(1). Should features in data be log10-transformed?

center

logical(1). Whether the variables should be shifted to be zero centered (\(\mu = 0\)).

scale

logical(1). Whether the variables should be scaled to have unit variance (\(\sigma = 1\)).

...

Optional arguments of the form variable = function() specifying the function to be applied to the specified column. Anonymous functions can be used but they should take a vector input and return a vector of equal length. See Examples. If also center/scaling, the ... transformation takes place before the center-scale step.

x

A rcp class object with instructions for pre-processing.

object

A recipe class object from the recipes package with the 3 pre-processing steps described above.

Value

A rcp class object containing information for the recipe steps.

Details

The order of recipe steps is always:

  1. log10-transform

  2. center

  3. scale

Functions

  • bake_recipe(): executes the recipe instructions defined during via create_recipe().

  • is.baked(): tests for presence of baked entry in attributes, indicating that the data have already been baked via bake_recipe().

  • convert_recipe(): converts an existing recipe object from the recipes package into a rcp object. Note that all conversions and not possible and the intended use case is for 3 step conversion:

    • log10-transform

    • centered

    • scaled

Examples

# create a pre-processing recipe
rcp <- create_recipe(mtcars)
rcp
#> ══ Pre-processing recipe ═════════════════════════════════════════════════
#> 
#> ─── Training data:
#> ✓ Data containing 32 samples used in recipe
#> ✓ RFU features ( n = 11 ) will be processed by:
#> 
#> ─── Steps:
#> • log10-transformed      ✓
#> • Centered (mean = 0)    ✓
#> • Scaled (sd = 1)        ✓
#> 
#> ══════════════════════════════════════════════════════════════════════════

rcp2 <- create_recipe(mtcars,
                      feat = c("mpg", "disp", "drat", "wt"),
                      disp = abs,
                      hp   = log10,
                      qsec = function(x) round(x / 10))

# apply recipe to orig/own data set
new_data <- bake_recipe(rcp, mtcars)

# Logical test
is.baked(new_data)
#> [1] TRUE

# converting recipes
library(recipes)
#> 
#> Attaching package: ‘recipes’
#> The following object is masked from ‘package:stats’:
#> 
#>     step
rec <- recipe(Species ~ ., data = iris) |>
  step_log(all_predictors(), base = 10) |>
  step_center(all_predictors()) |>
  step_scale(all_predictors()) |>
  prep()

convert_recipe(rec)
#> ══ Pre-processing recipe ═════════════════════════════════════════════════
#> 
#> ─── Training data:
#> ✓ Data containing 150 samples used in recipe
#> ✓ RFU features ( n = 4 ) will be processed by:
#> 
#> ─── Steps:
#> • log10-transformed      ✓
#> • Centered (mean = 0)    ✓
#> • Scaled (sd = 1)        ✓
#> 
#> ══════════════════════════════════════════════════════════════════════════