rrotsfspls: rotsf rrotsfspls

In whcsu/rotsf: Survival Forest with PCA and Partial Least Squares

Description

Grow a survival forest with partial least squares(PLS)

Usage

rrotsfspls(x, y, trlength = 100, mtry = floor(sqrt(ncol(x))),
  impute = TRUE, nplscomp = floor(sqrt(ncol(x))), vari_status = FALSE,
  ...)

Arguments

x	The covariates(predictor variables) of training data.
y	Survival time and censored status of training data. Must be a Surv survival object
trlength	The ensemle size (the number of base survival trees). Default is 100.
mtry	The number of covariates(predictor variables) used in each base tree model. Default is the square root of the number of all avaibable covariates.
vari_status	Whether or not calculate variables importance scores. Default is "FALSE".
...	Additional arguments for the base decision tree, see the rpart package for details.

Value

Object of class rrotsfspls with elements

pectrees	A list of base models pecRpart in pec R package of size trlength. To retrieve a particular base model: use pectrees[[i]], where i takes values between 1 and trlength
colindexes	A list of covaraite subspace index for each base tree.
trlength	Number of bases models trained.
rotms	A list of PLS weights matrix.To retrieve a particular weight matrix: use rotms[[i]], where i takes values between 1 and trlength
varimp	If vari_status=FALSE, return a Matrix of variable importance scores.

Author(s)

Hong Wang and Lifeng Zhou

References

• Zhou L, Xu Q, Wang H. (2015) Rotation survival forest for right censored data. PeerJ 3:e1009 https://doi.org/10.7717/peerj.1009.

• Zhou, L., Wang, H., & Xu, Q. (2016). Random rotation survival forest for high dimensional censored data. SpringerPlus, 5(1), 1425.

See Also

pec R package

Examples

set.seed(123)
require(rotsf)
require(survival)
## Survival Forest with PLS  with default settings
#Lung DATA
data(lung)
lung=na.omit(lung)
lung[,3]=lung[,3]-1
n=dim(lung)[1]
L=sample(1:n,ceiling(n*0.5))
trset<-lung[L,]
teset<-lung[-L,]
rii=c(2,3)
plssurvmodel=rrotsfspls(x=trset[,-rii],y=Surv(trset[,rii[1]], trset[,rii[2]]))
# Get the 1th base model
firstbasemodel=plssurvmodel$pectrees[[1]]
#second PLS weight matrix
secondweigmatrix=plssurvmodel$rotms[[2]]
plssurvmodel2=rrotsfspls(x=trset[,-rii],y=Surv(trset[,rii[1]], trset[,rii[2]]),vari_status=TRUE)
#variable importance
varimpscores=plssurvmodel2$varimp

whcsu/rotsf documentation built on Dec. 4, 2019, 2:10 p.m.