bin_manual: Bin the variables of a dataset manually

In jgendrinal/creditscore: Tidy workflows for credit scoring

Description

bin_manual cuts numeric, variables according to the mapping specified by the user. This is a standard step in feature engineering or data transformation for scorecard modelling.

Usage

bin_manual(.data, bad, ..., check = FALSE)

Arguments

.data	A tibble or data.frame
bad	A variable name indicating which rows are bad borrowers
...	Named lists separated by commas containing mappings of variables to a transformation. Cuts for each variable are separated by c_l or c_r.
check	A logical, if TRUE, bin_manual will check for a minimum of 30 borrowers for each bin. FALSE by default.

Details

Similar to the dplyr mutate, the function takes a named list of lists, separated by commas, with the variable name on the left side, and the transformation on the right side. bin_manual uses the c_l and c_r functions to close where the values of a variable will be cut and transformed into categories.

Value

A tibble or data.frame with the variables transformed into character vectors.

Examples

bin_manual(german, bad, age = c_l(40, 55))
bin_manual(german,
           bad,
           duration = c_r(15, 32),
           check = TRUE)

jgendrinal/creditscore documentation built on May 30, 2020, 3:51 p.m.