var.used.forestRK {forestRK}

Extract the list of covariates used to perform the splits to generate a particular tree(s) in a forestRK object

Description

Spits out the list of covariates used to perform the splits to generate a particular tree(s) in a forestRK object that the user provided.

The function extracts the list of names of covariates used in splits to construct a single or a multiple numbers of trees from a forestRK object. The var.used.forestRK displays the actual name of the covariate used for each split (not their numericized ones), consistent to the exact order of the split; for instance, the 1st element of the vector covariate.used.for.split.tree[["6"]] from the example below is the covariate on which the 1st split had occured while the 6th tree in the forestRK.1 object was built.

Each vector in the list are named by the exact indices of the tree; for example, if the code obj <- var.used.forestRK(forestRK.1, tree.index=c(4,5,6)) is used to extract the list of covariates used for splitting to construct 4th, 5th, and 6th trees in the forest, and the user can retrieve the information pertains explicitly to the 6th tree in the forest by doing obj[["6"]].

Usage

 var.used.forestRK(forestRK.object = forestRK(), tree.index = c())

Arguments

forestRK.object

a forestRK object.

tree.index

a vector storing the indices of the trees that we are interested to examine.

Value

A list of vectors that stores the names of covariates on which each split was performed to construct the specific tree(s) in a forestRK model that the user provided.

Author(s)

Hyunjin Cho, [e-mail] Rebecca Su, [e-mail]

See Also

forestRK

Examples

  library(forestRK)

  x.train <- x.organizer(iris[,1:4], encoding = "num")[c(1:25,51:75,101:125),]
  y.train <- y.organizer(iris[c(1:25,51:75,101:125),5])$y.new

  # random forest
  # min.num.obs.end.node.tree is set to 5 by default;
  # entropy is set to TRUE by default
  # normally nbags and samp.size have to be much larger than 30 and 50
  forestRK.1 <- forestRK(x.train, y.train, nbags = 30, samp.size = 50)

  # prediction from a random forest RK
  covariate.used.for.split.tree <- var.used.forestRK(forestRK.1,
                                                     tree.index=c(4,5,6))

  # retrieve the list of covariates used for splitting for the 'tree #6'
  covariate.used.for.split.tree[["6"]]