ci-assessment: Assessment of CI methods

In julian-gehring/Rariant: Identification and Assessment of Single Nucleotide Variants through Shifts in Non-Consensus Base Call Frequencies

Description

Functions to compute the coverage probability of a confidence interval method.

Usage

  coverageProbability(pars, fun = acCi, n_sample = 1e4, min_k, ...)

Arguments

pars	Data frame with parameter combinations [data.frame]
n_sample	Number of assessments per parameter combination [integer(1)].
fun	CI function
min_k	Minimum 'k2' value to use.
...	Additional arguments that are passed on to 'fun'.

Value

The 'data.frame' object 'pars' with additional columns 'cp' for the coverage probability and 'aw' average confidence interval width.

References

Fagerland, Morten W., Stian Lydersen, and Petter Laake. Recommended Confidence Intervals for Two Independent Binomial Proportions. Statistical Methods in Medical Research (2011).

Examples

## Define parameter space
pars = expand.grid(k1 = 1:5, k2 = 5, n1 = 30, n2 = 30)
conf_level = 0.95

## Compute coverage probabilities
cp = coverageProbability(pars, fun = acCi, n_sample = 1e2, conf_level = conf_level)
print(cp)

julian-gehring/Rariant documentation built on May 20, 2019, 4:20 a.m.