FDOTT (version 0.1.0)

tab_sample: Generate tabulated samples from probability vectors

Description

Generate count vectors instead of samples, i.e., vectors giving the number of times a sample was observed at the respective points.

Usage

tab_sample(n, mu, prob = FALSE)

Value

The count vectors corresponding to the generated samples. Has the same structure as mu.

Arguments

n

vector or nested list of sample sizes.

mu

matrix (row-wise) or nested list containing probability vectors to sample from. The structure of n and mu must be the same.

prob

logical value indicating whether probabilities (instead of counts) should be returned.

Examples

## matrix example

mu <- matrix(c(0.01, 0.99, 0.5, 0.5), 2, 2, TRUE)
n <- c(80, 20)

set.seed(123)
cv <- tab_sample(n, mu)
print(cv)
# sample sizes are rowsums
print(rowSums(cv))
# empirical probability vectors
print(sweep(cv, 1, n, "/"))
set.seed(123)
# same result
print(tab_sample(n, mu, prob = TRUE))

## list example

mu <- list(
   list(c(0.3, 0.7), c(0.25, 0.75)),
   list(c(0, 1), c(0.5, 0.5))
)
n <- list(list(100, 120), list(80, 90))

set.seed(123)
cv <- tab_sample(n, mu)
print(cv)
# empirical probability vectors
print(rapply(cv, \(x) x / sum(x), how = "replace"))
set.seed(123)
print(tab_sample(n, mu, prob = TRUE))