feamiR (version 0.1.0)

rfgini: Random Forest cumulative MeanDecreaseGini feature selection. Implements a feature selection approach based on cumulative MeanDecreaseGini using Random Forests trained on multiple subsamples.

Description

Random Forest cumulative MeanDecreaseGini feature selection. Implements a feature selection approach based on cumulative MeanDecreaseGini using Random Forests trained on multiple subsamples.

Usage

rfgini(num_runs = 100, num_trees = 30, file_path = file_path)

Arguments

num_runs

Number of subsamples to use for voting scheme (default: 100)

num_trees

Number of trees for random forest (selected using select_rf_numtrees)

file_path

Where the num_runs subsample files are found (e.g. if sample 10 is at 'subsamples/sample10.csv' then file_path should be 'subsamples/sample'). There must be enough samples to fulfill num_runs runs.

Value

The function will output a data.frame with cumulative mean decrease in Gini for each feature in the first columns (each row is a feature) and the rest of the column containing mean decrease in Gini for each of the num_runs runs.

Examples

# NOT RUN {
rfgini(
  num_runs=5,
  num_trees=30,
  file_path=paste(system.file('samples/subsamples',package = "feamiR"),'/sample',sep=''))
# }