iv_filter: Variable reduction based on Information Value filter

In scorecardModelUtils: Credit Scorecard Modelling Utils

Description

The function returns a list of variables that can be dropped because of low discriminatory power, based on Information Value. If IV for a variable is less than a user defined threshold, the variable will be recommended to be dropped by this function.

Usage

iv_filter(base, iv_table, threshold)

Arguments

base	input dataframe
iv_table	dataframe of class iv_table with two columns - Variable_name, iv
threshold	threshold IV value below which the variable will be recommended to be dropped

Value

An object of class "iv_filter" is a list containing the following components:

retain_var_tab	variables remaining post IV filter as a dataframe
retain_var_name	array of column names of variables to be retained
dropped_var_tab	variables that can be dropped based on IV filter as a dataframe
threshold	threshold IV value used as input parameter

Author(s)

Arya Poddar <[e-mail]>

Examples

data <- iris
data$Y <- sample(0:1,size=nrow(data),replace=TRUE)
x <- c("Sepal.Length","Sepal.Width","Petal.Length","Petal.Width")
iv_table_list <- iv_table(base = data,target = "Y",num_var_name = x,cat_var_name = "Species")
ivf_list <- iv_filter(base = data,iv_table = iv_table_list$iv_table,threshold = 0.02)
ivf_list$retain_var_tab
ivf_list$retain_var_name
ivf_list$dropped_var_tab
ivf_list$threshold

scorecardModelUtils documentation built on May 2, 2019, 9:59 a.m.