rfeControl: Controlling the Feature Selection Algorithms
In caret: Classification and Regression Training

rfeControl

R Documentation

Controlling the Feature Selection Algorithms

Description

This function generates a control object that can be used to specify the details of the feature selection algorithms used in this package.

Usage

rfeControl(
  functions = NULL,
  rerank = FALSE,
  method = "boot",
  saveDetails = FALSE,
  number = ifelse(method %in% c("cv", "repeatedcv"), 10, 25),
  repeats = ifelse(method %in% c("cv", "repeatedcv"), 1, number),
  verbose = FALSE,
  returnResamp = "final",
  p = 0.75,
  index = NULL,
  indexOut = NULL,
  timingSamps = 0,
  seeds = NA,
  allowParallel = TRUE
)

Arguments

`functions`	a list of functions for model fitting, prediction and variable importance (see Details below)
`rerank`	a logical: should variable importance be re-calculated each time features are removed?
`method`	The external resampling method: `boot`, `cv`, `LOOCV` or `LGOCV` (for repeated training/test splits
`saveDetails`	a logical to save the predictions and variable importances from the selection process
`number`	Either the number of folds or number of resampling iterations
`repeats`	For repeated k-fold cross-validation only: the number of complete sets of folds to compute
`verbose`	a logical to print a log for each external resampling iteration
`returnResamp`	A character string indicating how much of the resampled summary metrics should be saved. Values can be “final”, “all” or “none”
`p`	For leave-group out cross-validation: the training percentage
`index`	a list with elements for each external resampling iteration. Each list element is the sample rows used for training at that iteration.
`indexOut`	a list (the same length as `index`) that dictates which sample are held-out for each resample. If `NULL`, then the unique set of samples not contained in `index` is used.
`timingSamps`	the number of training set samples that will be used to measure the time for predicting samples (zero indicates that the prediction time should not be estimated).
`seeds`	an optional set of integers that will be used to set the seed at each resampling iteration. This is useful when the models are run in parallel. A value of `NA` will stop the seed from being set within the worker processes while a value of `NULL` will set the seeds using a random set of integers. Alternatively, a list can be used. The list should have `B+1` elements where `B` is the number of resamples. The first `B` elements of the list should be vectors of integers of length `P` where `P` is the number of subsets being evaluated (including the full set). The last element of the list only needs to be a single integer (for the final model). See the Examples section below.
`allowParallel`	if a parallel backend is loaded and available, should the function use it?

Details

More details on this function can be found at http://topepo.github.io/caret/recursive-feature-elimination.html#rfe.

Backwards selection requires function to be specified for some operations.

The fit function builds the model based on the current data set. The arguments for the function must be:

x the current training set of predictor data with the appropriate subset of variables
y the current outcome data (either a numeric or factor vector)
first a single logical value for whether the current predictor set has all possible variables
last similar to first, but TRUE when the last model is fit with the final subset size and predictors.
...optional arguments to pass to the fit function in the call to rfe

The function should return a model object that can be used to generate predictions.

The pred function returns a vector of predictions (numeric or factors) from the current model. The arguments are:

object the model generated by the fit function
x the current set of predictor set for the held-back samples

The rank function is used to return the predictors in the order of the most important to the least important. Inputs are:

object the model generated by the fit function
x the current set of predictor set for the training samples
y the current training outcomes

The function should return a data frame with a column called var that has the current variable names. The first row should be the most important predictor etc. Other columns can be included in the output and will be returned in the final rfe object.

The selectSize function determines the optimal number of predictors based on the resampling output. Inputs for the function are:

xa matrix with columns for the performance metrics and the number of variables, called "Variables"
metrica character string of the performance measure to optimize (e.g. "RMSE", "Rsquared", "Accuracy" or "Kappa")
maximizea single logical for whether the metric should be maximized

This function should return an integer corresponding to the optimal subset size. caret comes with two examples functions for this purpose: pickSizeBest and pickSizeTolerance.

After the optimal subset size is determined, the selectVar function will be used to calculate the best rankings for each variable across all the resampling iterations. Inputs for the function are:

y a list of variables importance for each resampling iteration and each subset size (generated by the user-defined rank function). In the example, each each of the cross-validation groups the output of the rank function is saved for each of the subset sizes (including the original subset). If the rankings are not recomputed at each iteration, the values will be the same within each cross-validation iteration.
size the integer returned by the selectSize function

This function should return a character string of predictor names (of length size) in the order of most important to least important

Examples of these functions are included in the package: lmFuncs, rfFuncs, treebagFuncs and nbFuncs.

Model details about these functions, including examples, are at http://topepo.github.io/caret/recursive-feature-elimination.html. .

Value

A list

Author(s)

Max Kuhn

Examples


  ## Not run: 
subsetSizes <- c(2, 4, 6, 8)
set.seed(123)
seeds <- vector(mode = "list", length = 51)
for(i in 1:50) seeds[[i]] <- sample.int(1000, length(subsetSizes) + 1)
seeds[[51]] <- sample.int(1000, 1)

set.seed(1)
rfMod <- rfe(bbbDescr, logBBB,
             sizes = subsetSizes,
             rfeControl = rfeControl(functions = rfFuncs,
                                     seeds = seeds,
                                     number = 50))
  
## End(Not run)

caret documentation built on March 31, 2023, 9:49 p.m.