hvp: Interaction term ranking with normalized high order...

In yifansha/highinthunt: High Order Interaction Variable Importance Through Random Forest

Description

This function ranks the interaction terms in a rfsrc object according to normalized high order interaction variable importance and compute other interaction variable importance.

Usage

hvp(joint, obj, importance = "permute",
    block = 1, choice = "hivimp")

Arguments

joint	A matrix that each row stores the variable combination in each interaction term.
obj	An object of class (rfsrc, grow).
importance	Method for computing variable importance (VIMP). It is the same as “importance" in rfsrc function in randomForestSRC package.
block	Specifies number of trees in a block when calculating VIMP. It is the same as “block.size" in vimp function in randomForestSRC package.
choice	Method(s) used for ranking interaction terms. Choose “hivimp" for high order interaction variable importance or/and choose “acuvimp" for exact high-order interaction deviance.

Value

A dataframe with ranking criteria for each interaction term in the row and several methods in the column. “joinvimp" for joint vimp; “HIvimp" for high order interaction variable importance; “nmHIvimp" for normalized high order interaction variable importance.

Author(s)

Yifan Sha and Min Lu

References

Ishwaran H. (2007). Variable importance in binary regression trees and forests, Electronic J. Statist., 1:519-537.

Ishwaran H., Kogalur U.B., Gorodeski E.Z, Minn A.J. and Lauer M.S. (2010). High-dimensional variable selection for survival data. J. Amer. Statist. Assoc., 105:205-217.

Ishwaran H., Kogalur U.B., Chen X. and Minn A.J. (2011). Random survival forests for high-dimensional data. Statist. Anal. Data Mining, 4:115-132.

Examples

data(express)
o.0 <- rfsrc(y~., data = express[,1:11])
joint <- t(combn(1:length(o.0$xvar.names),3))
o <- hvp(joint, obj = o.0, importance = "permute",
         block = 1, choice = c("hivimp", "acuvimp"))
head(o)

yifansha/highinthunt documentation built on July 2, 2020, 6:29 p.m.