R/step_select_forests.R
step_select_forests.Rd`step_select_forests` creates a *specification* of a recipe step that selects a subset of predictors based on the ranking of variable importance using a `parsnip::rand_forest` supported model.
step_select_forests(
recipe,
...,
outcome = NULL,
role = "predictor",
trained = FALSE,
engine = "ranger",
options = list(importance = "permutation"),
mtry = NULL,
trees = NULL,
min_n = NULL,
top_p = NA,
threshold = NA,
exclude = NULL,
scores = NULL,
skip = FALSE,
id = recipes::rand_id("select_forests")
)
# S3 method for step_select_forests
tidy(x, ...)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.
A character string with the name of the response variable to use to calculate the feature importance scores.
Not used by this step since no new variables are created.
A logical to indicate if the quantities for preprocessing have been estimated.
A supported rand_forest engine that is supported by parsnip. The default is "ranger".
A named list of options to pass to the rand_forest engine. For example, if `engine = 'ranger'` (the default) then options could be `list(permutation = 'importance`) because a feature importance method needs to be specified for this engine. This is the default.
An integer for the number of predictors that will be randomly sampled at each split when creating the tree models.
An integer for the number of trees contained in the ensemble.
An integer for the minimum number of data points in a node that are required for the node to be split further.
An integer with the number of best scoring features to select.
A numeric value between 0 and 1 representing the percentile of best scoring features to select. Features with scores that are _larger_ than the specified threshold will be retained, for example `threshold = 0.9` will retain only predictors with scores in the top 90th percentile. Note that this overrides `top_p`.
A character vector of predictor names that will be removed from the data. This will be set when `prep()` is used on the recipe and should not be set by the user.
A tibble with 'variable' and 'scores' columns containing the names of the variables and their feature importance scores. This parameter is only produced after the recipe has been trained.
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.
A character string that is unique to this step to identify it.
A `step_select_forests` object.
a `step_select_forests` object.
library(recipes)
library(parsnip)
# load the example iris dataset
data(cells, package = "modeldata")
# create a preprocessing recipe
rec <-
recipe(class ~ ., data = cells[, -1]) %>%
step_select_forests(all_predictors(), outcome = "class", top_p = 10,
threshold = 0.9)
prepped <- prep(rec)
preproc_data <- juice(prepped)
prepped
#> Recipe
#>
#> Inputs:
#>
#> role #variables
#> outcome 1
#> predictor 56
#>
#> Training data contained 2019 data points and no missing data.
#>
#> Operations:
#>
#> Variable importance feature selection (50 excluded)