Estimate error probability, w

Note that tol = .Machine$double.eps unless overwritten.

Usage

estimate_w(x, use_mpfr = FALSE, ...)

Arguments

x

either a 3x3 contingency table or 3-vector with counts (n1, n2, n3) where n1 = sum(diag(x)), n2 = x[1L, 3L] + x[3L, 1L], n3 = sum(x) - n1 - n2

use_mpfr

whether to use Rmpfr::optimizeR()] for arbitrary precision

...

passed on to optimise() or to Rmpfr::optimizeR if use_mpfr = TRUE

Examples

tab1 <- matrix(c(1000, 10, 2, 12, 100, 8, 1, 7, 200), nrow = 3)
tab1
#>      [,1] [,2] [,3]
#> [1,] 1000   12    1
#> [2,]   10  100    7
#> [3,]    2    8  200
estimate_w(tab1)
#> [1] 0.008154441
n1 <- sum(diag(tab1))
n2 <- tab1[1L, 3L] + tab1[3L, 1L]
n3 <- sum(tab1) - n1 - n2
estimate_w(c(n1, n2, n3))
#> [1] 0.008154441

tab2 <- structure(c(40000L, 30L, 5L, 15L, 400L, 2L, 5L, 5L, 350L),
        dim = c(3L, 3L), class = "table")
tab2
#>      [,1]  [,2]  [,3] 
#> [1,] 40000    15     5
#> [2,]    30   400     5
#> [3,]     5     2   350
estimate_w(tab2)
#> [1] 0.0004414364
estimate_w(tab2, use_mpfr = TRUE)
#> 
#> Attaching package: ‘gmp’
#> The following objects are masked from ‘package:base’:
#> 
#>     %*%, apply, crossprod, matrix, tcrossprod
#> C code of R package 'Rmpfr': GMP using 64 bits per limb
#> 
#> Attaching package: ‘Rmpfr’
#> The following object is masked from ‘package:gmp’:
#> 
#>     outer
#> The following objects are masked from ‘package:stats’:
#> 
#>     dbinom, dgamma, dnbinom, dnorm, dpois, dt, pnorm
#> The following objects are masked from ‘package:base’:
#> 
#>     cbind, pmax, pmin, rbind
#> 1 'mpfr' number of precision  132   bits 
#> [1] 0.00044143635878284077344503195817562240248647