Skip to contents

c_CV: Coefficient of Variation (CV) of false discovery rate (c).

Source: R/callDensityEstimate.R
c_CV.Rd

CV.c is calculated from the variance of c, which depends on the probability of false alarm (c), a proportion.

Usage

c_CV(n, c)

Arguments

n

  • Number of trials used to estimate c

c

  • false discovery rate (proportion of false positive detections divided by the total number predicted positive)

Value

cv.c - coefficient of variation of the false discovery rate

Details

Rationale for this calculation Variance of a proportion, c derived from a binomial process. Zar 2010 Eqn. 24.17 indicates sigma^2 = pq/(n-1) here p = c, q = (1-c), and n is the number of detections inspected when estimating c, so:

  1. var(c)= c * (1-c)/(n-1)

And required inputs are n, the number of trials conducted (i.e. how many detections were inspected when estimating c) and the estimate of c (as a proportion between 0 and 1 inclusive).