CorReg (version 1.0.5)

recursive_tree: decision tree in a recursive way

Description

decision tree in a recursive way

Usage

recursive_tree(data = data, Y = "Y", modele = NULL, kill = NULL,
  index = NULL, print = TRUE, plot = TRUE, main = NULL, sub = NULL,
  lang = c("en", "fr"))

Arguments

data

the dataset including the response

Y

the name of the response

modele

(optional) vector of names of covariates allowed in the tree

kill

vector of the names to kill (variables won't be used in the tree)

index

to give a number to the plot

print

boolean to print the tree parameters

plot

boolean to plot the tree

main

the main title if plot=TRUE

sub

the subtitle (if NULL it is automatically added)

lang

the language for the automatic subtitle in the plot

Value

• returns the tree as an "rpart" object and the modele as a vector of the names of the covariates the tree could have used (to give as an input of the function).
• modele
  vector of the names of the covariates the tree could have used
• tree
  the regression tree as an "rpart" object

Examples

data<-mtcars
   require(CorReg)
main="Regression tree of cars consumption (in mpg)"
  mytree=recursive_tree(data = data,Y ="mpg" ,main=main)
   #want to try without cylinder and disp
  mytree2=recursive_tree(data = data,Y ="mpg" ,kill=c("cyl","disp"),modele=mytree$modele,main=main)