Skip to main content

Table 1 Method overview

From: EFS: an ensemble feature selection tool implemented as R-package and web-application

Command

Parameters

Information

ensemble_fs

data

object of class data.frame

 

classnumber

index of variable for binary classification

 

NA_threshold

threshold for deletion of features with a greater proportion of NAs

 

cor_threshold

correlation threshold within features

 

runs

amount of runs for randomForest and cforest

 

selection

selection of feature selection methods to be conducted

barplot_fs

name

character string giving the name of the file

 

efs_table

table object of class matrix retrieved from ensemble_fs

efs_eval

data

object of class data.frame

 

efs_table

table object of class matrix retrieved from ensemble_fs

 

file_name

character string, name which is used for the two possible PDF files.

 

classnumber

index of variable for binary classification

 

NA_threshold

threshold for deletion of features with a greater proportion of NAs

 

logreg

logical value indicating whether to conduct an evaluation via logistic regression or not

 

permutation

logical value indicating whether to conduct a permutation of the class variable or not

 

p_num

number of permutations; default set to a 100

 

variances

logical value indicating whether to calculate the variances of importances retrieved

  

from bootstrapping or not

 

jaccard

logical value indicating whether to calculate the Jaccard-index or not

 

bs_num

number of bootstrap permutations of the importances

 

bs_percentage

proportion of randomly selected samples for bootstrapping

  1. The R-package EFS provides three functions