find_n_ksigma: Find N for k-sigma sample interval

In dygobeng/nterval: Calculate Sample Size for Population Coverage Intervals

Find N for k-sigma sample interval

Description

Find required sample size for "k-sigma sample interval" (i.e., parametric sample interval applied to a Normal distribution) to achieve targeted reliability using bisection algorithm

Usage

find_n_ksigma(
  proximity_range,
  reliability,
  k = NULL,
  search_interval = c(3, 500),
  n_sim = 1e+06,
  tolerance = 1e-05,
  plot = FALSE,
  verbose = FALSE,
  seed = NULL
)

Arguments

proximity_range	Range within which sample coverage should fall with pre-specified reliability.
reliability	Targeted probability that individual coverage of parametric sample interval applied to a Normal distribution falls within proximity_range.
k	Standard deviation multiplier for parametric sample interval applied to Normal distribution (typically 2 or 3). If NULL, k is assigned based on the midpoint of the proximity_range.
search_interval	Initial interval of sample sizes for bisection algorithm. Default is c(3, 500).
n_sim	Number of simulation iterations used to estimate reliability. Default is 1000000.
tolerance	Threshold for difference between targeted and estimated reliability that indicates sufficient proximity of the root-finding function to 0, triggering the end of the search. Default is 0.00001.
plot	If TRUE, plot sampling distribution for true interval coverage using sample size identified through bisection.
verbose	If TRUE, print the search interval used for each round of the bisection algorithm search.
seed	Randomization seed. Defaults to NULL so that no seed is used.

Value

A list with the following elements:

• The estimated sample size required to achieve target reliability.

• If plot is TRUE, the plot object.

Examples

find_n_ksigma(
  proximity_range = c(0.93, 0.97),
  reliability = 0.7,
  k = 2
)

dygobeng/nterval documentation built on March 22, 2022, 6:40 p.m.