compute.threshold.pooledROC.BB: Pooled ROC-based threshold values.

In AROC: Covariate-Adjusted Receiver Operating Characteristic Curve Inference

Pooled ROC-based threshold values.

Description

Estimates pooled ROC-based threshold values using the Bayesian bootstrap estimator proposed by Gu et al. (2008).

Usage

compute.threshold.pooledROC.BB(object, FPF = 0.5)

Arguments

object	An object of class AROC as produced by pooledROC.BB.
FPF	Numeric vector with the FPF at which to calculate the pooled ROC-based threshold values. Atomic values are also valid.

Value

As a result, the function provides a list with the following components:

thresholds	A matrix with the posterior mean and posterior 2.5% and 97.5% quantiles of the pooled ROC-based threshold values. The matrix has as many rows as different FPFs.
FPF	the supplied FPF argument
TPF	TPFs corresponding to the estimated threshold. In addition to the posterior mean, the 95% pointwise credible band is also returned.

References

Gu, J., Ghosal, S., and Roy, A. (2008). Bayesian bootstrap estimation of ROC curve. Statistics in Medicine, 27, 5407–5420.

See Also

pooledROC.BB

Examples

library(AROC)
data(psa)
# Select the last measurement
newpsa <- psa[!duplicated(psa$id, fromLast = TRUE),]

# Log-transform the biomarker
newpsa$l_marker1 <- log(newpsa$marker1)


m0_BB <- pooledROC.BB(y0 = newpsa$l_marker1[newpsa$status == 0], 
            y1 = newpsa$l_marker1[newpsa$status == 1], p = seq(0,1,l=101), B = 5000)

### Threshold values for a fixed FPF
th_m0_BB <- compute.threshold.pooledROC.BB(m0_BB, FPF = 0.1)

th_m0_BB$threshold

AROC documentation built on March 18, 2022, 5:22 p.m.