Skip to contents

Variable selection

Source: R/orsf_vs.R
orsf_vs.Rd

Variable selection

Usage

orsf_vs(object, n_predictor_min = 3, verbose_progress = NULL)

Arguments

object

(ObliqueForest) a trained oblique random forest object (see orsf).

n_predictor_min

(integer) the minimum number of predictors allowed

verbose_progress

(logical) not implemented yet. Should progress be printed to the console?

Value

a data.table with four columns:

  • n_predictors: the number of predictors used

  • stat_value: the out-of-bag statistic

  • predictors_included: the names of the predictors included

  • predictor_dropped: the predictor selected to be dropped

Details

tree_seeds should be specified in object so that each successive run of orsf will be evaluated in the same out-of-bag samples as the initial run.

Examples


object <- orsf(formula = time + status ~ .,
               data = pbc_orsf,
               n_tree = 25,
               importance = 'anova')

orsf_vs(object, n_predictor_min = 15)
#>    n_predictors stat_value                        variables_included
#>           <int>      <num>                                    <list>
#> 1:           15  0.8356685     age,albumin,ascites,ast,bili,chol,...
#> 2:           16  0.8351997     age,albumin,ascites,ast,bili,chol,...
#> 3:           17  0.8296786     age,albumin,ascites,ast,bili,chol,...
#> 4:           18  0.8185322 age,albumin,alk.phos,ascites,ast,bili,...
#>                                predictors_included predictor_dropped
#>                                             <list>            <char>
#> 1:  id,age,sex_f,ascites_1,spiders_1,edema_0.5,...          platelet
#> 2:   id,age,sex_f,ascites_1,hepato_1,spiders_1,...          hepato_1
#> 3: id,trt_placebo,age,sex_f,ascites_1,hepato_1,...       trt_placebo
#> 4: id,trt_placebo,age,sex_f,ascites_1,hepato_1,...          alk.phos