generate_data: Generate data following different missingness mechanisms

In adimajo/scoring: Credit Scoring Tools

Generate data following different missingness mechanisms

Description

This function performs generates

Usage

generate_data(n = 100, d = 3, type = "MAR well specified")

Arguments

n	The number of samples to return.
d	The dimension of samples to return.
type	The matrix of financed clients' labels

Details

This function generates data from a uniform(0,1) distribution, and generates labels y according to a logistic regression on this data with random -1/1 parameter for each coordinate (MAR well-specified), the square of this data (MAR misspecified), or this data and some additional feature (from U(0,1) as well - MNAR).

Value

Dataframe containing features as x.1..d, labels as y.

Author(s)

Adrien Ehrhardt

References

Ehrhardt, A., Biernacki, C., Vandewalle, V., Heinrich, P. and Beben, S. (2018), Reject Inference Methods in Credit Scoring: a rational review,

Examples

# We simulate data from financed clients
generate_data(n = 100, d = 3, type = "MAR well specified")

adimajo/scoring documentation built on March 7, 2024, 11:18 p.m.