runHSS: runHSS

runHSSR Documentation

runHSS

Description

This function runs hybrid subset selection.

Usage

runHSS(x, y, score.method, custom.score.method = NULL, downsample = TRUE)

Arguments

x

table with predictors of interest

y

vector of class labels

score.method

scoring metric for feature selection using HSS. Options include: 'euclidean', 'silhouette', 'pixel.density', 'pixel.entropy', or 'custom'.

custom.score.method

function for your custom scoring metric. Score.method must be 'custom'

downsample

boolean indicating whether to downsample data. Defaults to TRUE. Downsamples to 5000 cells per class label, or balanced sampling based on the class label with minimum # of cells.


mamouzgar/hsslda documentation built on March 20, 2023, 6 a.m.