sc.Misc: Miscellaneous tools for Credit Scoring

In lmtleminh/easysc: General Scorecard buiding functions

Description

These are set of tools for easy credit scorecard build. The set includes sc.score, sc.point and sc.corplot.

Usage

sc.score(data, model, pdo, score, odd)

sc.point(data, woe, model, pdo, score, odd)

sc.corplot(model, data, s = NULL)

sc.performance(model, X, y, s = NULL, plot = F)

Arguments

data	A data frame which contains target varible as well as predictor variables.
model	A model built.
pdo	Point double odds.
score	Score at which the desired odd is.
odd	The desired odd.
woe	A woe result which is calculated by woe.
s	In case of using Lasso model. Lambda min or lambda at 1se is chosen.
X	A data frame which has only predictor variables.
y	A vector of target variable.
plot	A logical scalar. Plot ROC curve.

Value

sc.score. The output is a data frame with chosen variables in the model. All variables are in score point. There is also a final score.

sc.point. The output is a list with chosen variables in the model. Variables with their values and score point.

sc.corplot. The output plot is generated showing correlated variables.

sc.performance. The output GINI & KS is printed.

Examples

## Not run: 
# Load library
library(easysc)

# Generate final score data based on the model built with data.
scoredata <- sc.score(data = df, model = model.Lasso, pdo = 100, score = 800, odd = 5)

# Generate final scorecard based on the model built with data.
scorecard <- sc.point(data = df, woe = woe, model = model.Lasso, pdo = 100, score = 800, odd = 5)

## End(Not run)

lmtleminh/easysc documentation built on July 5, 2019, 11:48 a.m.