R/CIDists.R
hyperCI.Rd
Computes a confidence interval for population size (N) in hypergeometric distribution.
hyperCI(M, n, m, conf.level = 0.95)
M | Number of successes in the population. |
---|---|
n | Number of observations in the sample. |
m | Number of observed successes in the sample. |
conf.level | Level of confidence to use for constructing confidence intervals (default is |
A 1x2 matrix that contains the lower and upper confidence interval bounds.
This is an inefficient brute-force algorithm. The algorithm computes the conf.level
range of possible values for m
, as if it was unknown, for a large range of values of N. It then finds all possible values of N for which m
was in the conf.level
range. The smallest and largest values of N for which m
was in the conf.level
range are the CI endpoints.
This algorithm is experimental at this point.
hyperCI(50,25,10)#> 95% LCI 95% UCI #> [1,] 86 228