create.biased.idx: Create biased indices for testing

In zellerlab/SIMBA: Simulation of Microbiome Data with Biological Accuracy

Create biased indices for testing

Description

This function creates biased indices for testing with an additional confounder

Usage

create.biased.idx(sim.location, subsets, repetitions,
lab.svar = NULL, meta.svar = NULL, strat.scheme = 'all', bias = 0.5)

Arguments

sim.location	file name for the .h5 file containing the simulations
subsets	vector of intended subset sizes for resampled data
repetitions	number of unique sets of indices to generate for each subset size deemed valid
lab.svar	for real data, name of the main stratification "label" variable in the original metadata; defaults to NULL and is not applicable for simulated data
meta.svar	for real data, a single string or vector of strings corresponding to the name(s) of binary metadata variable(s) by which biased resampling should be stratified; defaults to NULL and is not applicable for simulated data
strat.scheme	character, name of the index group to be written within the .h5 file containing different indices for downstream testing Can be either 'biomarker','confounder', 'random', or 'all' (the default)
bias	float, strength of the confounder bias in the subsampling

Details

todo

Value

vector of real metadata variables insufficient for resampling, if applicable, otherwise nothing returned & indices are written to 'sim.location'

zellerlab/SIMBA documentation built on June 2, 2025, 6:25 a.m.