Rprofet (version 3.1.1)

BinProfet: Coarse Binning Variable(s)

Description

Function that bins selected variable(s) and returns a dataframe with binned values. Uses greedy binning algorithm to perform coarse binning of selected variable(s).

Usage

BinProfet(
  data,
  id,
  target,
  varcol,
  min.cat = 4,
  num.bins = 10,
  min.pts.bin = 25,
  bracket = "left",
  special.values = NULL,
  sort_id = FALSE
)

Value

A dataframe containing the ID, target, and binned variable(s) with corresponding binned values.

Arguments

data

Dataframe of that contains ID, binary target and variables to be binned.

id

ID variable. See 'Details'.

target

The binary target/response variable for WOE. See 'Details'.

varcol

Vector of variables to be binned.

min.cat

Minimum number of bins.

num.bins

Target number of bins. Overridden by the number of levels if varcol is factor.

min.pts.bin

Minimum number of observations in a bin.

bracket

Indicating if the intervals should be closed on the right or left. Options include left and right.

special.values

A vector of values that should have their own bin. See 'Details'.

sort_id

Logical. The default is FALSE which does not sort the data by ID column. If TRUE, then data is sorted increasingly by ID column.

Examples

mydata <- ISLR::Default
head(mydata)
mydata$ID <- seq(1:nrow(mydata)) ## make an ID variable
mydata$default <- ifelse(mydata$default=="Yes", 1, 0) ## target coded with 1, 0

## bin balance and income
binned1 <- BinProfet(mydata, id="ID", target="default",
                        varcol = c("balance",  "income"), num.bins = 5)
head(binned1)

## bin categorical variable-------------------
binned2 <- BinProfet(mydata, id="ID", target="default",
                        varcol = "student", num.bins = 5)
head(binned2)
summary(binned2$student_Bins) ## num.bins overriden