Create class-balanced case-control groups for classification models

make_groups() creates class-balanced case-control groups for classification models. It separates the data into control and case groups. It then calculates the amount of data from each case in the control group and randomly selects the number of each control class samples. It also filters the control data of sex specific cases.

Usage

make_groups(
  join_data,
  variable = "Disease",
  case,
  cases,
  only_female = NULL,
  only_male = NULL,
  seed = 123
)

Arguments

join_data: Olink data in wide format joined with metadata.
variable: The variable to predict. Default is "Disease".
case: Case to predict.
cases: Cases.
only_female: Cases that are female specific.
only_male: Cases that are male specific.
seed: Seed for reproducibility. Default is 123.

Value

A list with combined, class-balanced control-case groups for each case.