filter_control: Control for filtering step of 5-STAR algorithm

View source: R/5STARcorefun.R

filter_controlR Documentation

Control for filtering step of 5-STAR algorithm

Description

Parameters for control of the elastic net or random forest filtering step (step 2 of the 5-STAR algorithm)

Usage

filter_control(method = "ENET", lambdatype = "min", mixparm = NULL,
  vimpalpha = 0.05, nfolds = 10, filterseed = 2019, ...)

Arguments

method

specifying method used for filtering; either "ENET" for elastic net, "RF" for random forest using defaults in randomForestSRC package (with possible modifications through "..."), or "RFbest" for tuned random forest (slower but more optimized)

lambdatype

Optional elastic net parameter; whether to use lambda that minimizes cross validation error (lambdatype="min", default) or the largest lambda that gives error within 1 standard error of minimum error (lambdatype="1se"). Ignored when method is "RF" or "RFbest"

mixparm

Optional elastic net mixing parameter alpha or grid of alpha values to search over. If nothing is entered, will search for best value between 0.05 and 0.95. Ignored when method is "RF" or "RFbest"

vimpalpha

For "RF" or "RFbest", the significance level for variable importance (VIMP) confidence intervals (CIs) to determine which covariates are passed through the filtering stage (default = 0.05, ignored for method == "ENET")

nfolds

Number of folds used for cross validation when tuning the elastic net tuning parameter lambda (Default = 10; ignored when method is "RF" or "RFbest")

filterseed

Seed for the filtering step to control variability in cross validation step of the elastic net filtering

...

Optional additional arguments passed into glmnet or rfsrc functions

Value

A list of control parameters for filtering step


rmarceauwest/fiveSTAR documentation built on June 30, 2023, 7:38 a.m.