R/uci.binomial.R

In davidavdav/sretools.R: Compute error metrics and make DET plots for NIST-like Speaker Recognition Evaluations

`uci.binomial` <-
function(x,n,a=0,b=1,alpha=.05){
  ## Finds the exact upper confidence limit for binomial count
  ## R v0.90.1, Tomas Aragon, created 01/01/2000, edited
  ## x = observed count (non-negative integer)
  ## n = number of Bernoulli trials
  ## a = default lower bound for bisection method
  ## b = default upper bound for bisection method
  ## alpha = .05 (default), for 95% confidence interval
  ## This function is used by ci.binomial()
  
  if(x==0) answer <- 1-(alpha)^(1/n)
  else{
    answer <- (a+b)/2
    if(abs(a-b) >= 0.00000001){
      lefta <- pbinom(x,n,a)-alpha/2
      centera <- pbinom(x,n,answer)-alpha/2
      righta <- pbinom(x,n,b)-alpha/2
      if(lefta*centera < 0){
        answer <- uci.binomial(x,n,a,answer)
      }
      else{
        answer <- uci.binomial(x,n,answer,b)
      }
    }
  }
  answer
}