compute_fdr_tpr: Compute the achieved FDR and TPR

View source: R/bayesian_cure_rate_model.R

compute_fdr_tprR Documentation

Compute the achieved FDR and TPR

Description

This help function computes the achieved False Discovery Rate (FDR) and True Positive Rate (TPR). Useful for simulation studies where the ground truth classification of subjects in susceptibles and cured items is known.

Usage

compute_fdr_tpr(true_latent_status, posterior_probs, 
	myCut = c(0.01, 0.05, 0.1, 0.15))

Arguments

true_latent_status

a binary vector containing the true latent status: 1 should correspond to the positive instances ("cured") and 0 to the negative ("susceptibles").

posterior_probs

a numeric vector with entries between 0 and 1 containing the scores (posterior probabilities) of being positive ("cured") for each item.

myCut

Vector containing the desired nominal FDR levels.

Value

This function will return for every nominal FDR level the following quantities:

achieved_fdr

the achieved false discovery rate.

tpr

the true positive rate.

nominal_fdr

the nominal FDR level.

Author(s)

Panagiotis Papastamoulis

Examples

set.seed(1)
v1 <- sample(0:1, size = 100, replace=TRUE, prob=c(0.8,0.2) )
v2 <- runif(100)
compute_fdr_tpr(true_latent_status = v1, posterior_probs = v2)

bayesCureRateModel documentation built on Oct. 4, 2024, 1:07 a.m.