Apply feature selection for the input data

Source: R/feature_selection.R

Take input as predictons and response and return the list of features that are important

feature_selection(X, y, mode, n_features)

Arguments

X

data.frame of predictors
y

vector of response and should be a factor in case of classification
mode

string regression or classification
n_features

int number of top important features select from all the features

Value

vector of feature nams of top n_features

Examples

feature_selection(data.frame(X1 = c(2, 4, 3), X2 = c(8, 7, 4)), c(4, 3, 5), "regression", 2)
#> Warning: package 'caret' was built under R version 3.6.2
#> Loading required package: lattice
#> Loading required package: ggplot2
#> Warning: prediction from a rank-deficient fit may be misleading
#> Warning: prediction from a rank-deficient fit may be misleading
#> Warning: prediction from a rank-deficient fit may be misleading
#> [1] X1 X2
#> Levels: X1 X2
feature_selection(data.frame(X1 = c(2, 4, 3), X2 = c(8, 7, 4)), factor(c(1, 1, 0)), "classification", 1)
#> Warning: prediction from a rank-deficient fit may be misleading
#> Warning: rfe is expecting 2 importance values but only has 1. This may be caused by having zero-variance predictors, excessively-correlated predictors, factor predictors that were expanded into dummy variables or you may have failed to drop one of your dummy variables.
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> Warning: prediction from a rank-deficient fit may be misleading
#> Warning: rfe is expecting 2 importance values but only has 1. This may be caused by having zero-variance predictors, excessively-correlated predictors, factor predictors that were expanded into dummy variables or you may have failed to drop one of your dummy variables.
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> Warning: prediction from a rank-deficient fit may be misleading
#> Warning: rfe is expecting 2 importance values but only has 1. This may be caused by having zero-variance predictors, excessively-correlated predictors, factor predictors that were expanded into dummy variables or you may have failed to drop one of your dummy variables.
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> Warning: There were missing importance values. There may be linear dependencies in your predictor variables
#> [1] "X1"