rfInterp: Variable Selection with Random Forest

In dustinfife/fifer: A Biostatisticians Toolbox for Various Activities, Including Plotting, Data Cleanup, and Data Analysis

Description

Find best variables using Random Forest (Interpretation Step). Find complete documentation at rfThresh

Usage

rfInterp(object, nruns = 20, nsd = 1, importance = "permutation", ...)

Arguments

object	a rfThresh object
nruns	how many forests should be grown?
nsd	defaults to one.
importance	method of calculating importance (permutation or gini)
...	other arguments passed to cforest or randomForest

Value

varselect.interp	The variables selected for Interpretation (sorted)
err.interp	The error at each stage of the stepwise variable inclusion.
sd.min	The standard deviation of the minimum fitted value.
num.varselect.interp	The number of variables selected for interpretation.
comput.time	Computation time of the procedure.
data	The dataset used for fitting the RF algorithm
formula	The formula of all variables included after the interpretation step.

Author(s)

Robin Genuer, Jean-Michel Poggi and Christine Tuleau-Malot, with modifications by Dustin Fife

See Also

rfInterp, rfPred

Examples

	#### do threshold step
## Not run: data(iris); 
data = iris; 
formula = as.formula("Species ~ Sepal.Length + Sepal.Width + Petal.Length + Petal.Width")
object = rfThresh(formula, data=iris, nruns=2, importance="gini"); 
#### run interpretation step
rfInterp(object, nruns=10, importance="permutation")
## End(Not run)

dustinfife/fifer documentation built on Oct. 31, 2020, 3:36 p.m.