imputeData: imputeData

In sontron/madis: Manipulating and Analyzing Data Interactively with Shiny

Description

using tree model to impute na values in a data

Usage

imputeData(data, impVars, modelVars, seed = 1, treeMethod = c("rpart",
  "ctree"), treeParams = list(), method = c("sample", "pred")[1])

Arguments

data	data.frame
impVars	variables needed imputed
modelVars	variables used in cart tree model
treeParams	parameters control aspects of cart tree.
method	impute methods, could be sample and pred. when using sample, the na values will be imputed from the raw data in the terminal node of a tree model. when using pred, the na values will be imputed by the predicted value of cart tree model.

Examples

mt<-mtcars
set.seed(123)
mt[cbind(sample(1:30,10,rep=F),sample(1:3,10,rep=T))]<-NA
imputeData(mt,impVars=names(mt)[1:3],modelVars=names(mt),treeParams=list(maxdepth=5,minbucket=7,minsplit=20,cp=0.01),method='sample')->mtImpSample
imputeData(mt,impVars=names(mt)[1:3],modelVars=names(mt),treeParams=list(maxdepth=5,minbucket=7,minsplit=20,cp=0.01),method='pred')->mtImpPred

plot(mtcars$mpg,mtImpSample$mpg)
plot(mtcars$mpg,mtImpPred$mpg)

sontron/madis documentation built on March 23, 2021, 10:17 p.m.