undersample_mindist: Undersample a dataset by iteratively removing the observation...

In scutr: Balancing Multiclass Datasets for Classification Tasks

Undersample a dataset by iteratively removing the observation with the lowest total distance to its neighbors of the same class.

Description

Undersample a dataset by iteratively removing the observation with the lowest total distance to its neighbors of the same class.

Usage

undersample_mindist(data, cls, cls_col, m, ...)

Arguments

data	Dataset to undersample. Aside from cls_col, must be numeric.
cls	Class to be undersampled.
cls_col	Column containing class information.
m	Desired number of observations after undersampling.
...	Additional arguments passed to dist().

Value

An undersampled dataframe.

Examples

setosa <- iris[iris$Species == "setosa", ]
nrow(setosa)
undersamp <- undersample_mindist(setosa, "setosa", "Species", 50)
nrow(undersamp)

scutr documentation built on Nov. 18, 2023, 1:08 a.m.