mi_balance_data {MantaID}

Data balance. Most classes adopt random undersampling, while a few classes adopt smote method to oversample to obtain relatively balanced data;

Description

Data balance. Most classes adopt random undersampling, while a few classes adopt smote method to oversample to obtain relatively balanced data;

Usage

mi_balance_data(data, ratio = 0.3, parallel = FALSE)

Arguments

data

A data frame. Except class column, all are numeric types.

ratio

Numeric between 0 and 1. The percent of test set split from data.

parallel

Logical.

Value

A list contain train set and test set.

Examples

library(dplyr)
data = rename(iris,class =Species)
mi_balance_data(data)