undersample_mclust: Undersample a dataset by expectation-maximization clustering

In scutr: Balancing Multiclass Datasets for Classification Tasks

Description

Undersample a dataset by expectation-maximization clustering

Usage

undersample_mclust(data, cls, cls_col, m)

Arguments

data	Data to be undersampled.
cls	Class to be undersampled.
cls_col	Class column.
m	Number of samples in undersampled dataset.

Value

The undersampled dataframe containing only instance of cls.

Examples

setosa <- iris[iris$Species == "setosa", ]
nrow(setosa)
undersamp <- undersample_mclust(setosa, "setosa", "Species", 15)
nrow(undersamp)

scutr documentation built on June 24, 2021, 5:07 p.m.