Sample using a multinomial distribution — sample

Sample from a multinomial distribution given a vector of counts or proportions and a desired number of specimens in the returned sample.

Usage

sample_multinomial(x, sample_size)

Arguments

x: A numeric vector of counts per bin. The values do not need to be integer values or sum to one because they will be converted to proportions using proportions() prior to being passed to the prop argument of rmultinom().
sample_size: A single, positive integer specifying the number of samples you want to partition into bins. For example, if sample_size = 20 it is assumed that you want 20 fish in your returned sample partitioned into bins. There is no default.

Value

A vector the same length as x is returned, where the sum of the vector equals the input to sample_size.

Details

Sampling from a multinomial distribution is often used when composition data are simulated. For example, in your operating model you will know how many fish there are per age bin. The number of fish per age bin is transformed into proportions per age bin and the user can specify the final sample size of fish that they want in the returned vector.

The number of fish per age bin often include fractional fish in a simulation because numbers of fish will be calculated from biomass and mean weight at age. Thus, the input vector to x can be integers, between 0 and 1, or real numbers. Everything will work but the result will always be integers representing the number of fish per bin.

Examples

# Sample from a multinomial distribution with 20 fish and 3 age bins
set.seed(123)
sample_multinomial(c(10, 5, 5), sample_size = 20)
#> [1] 9 7 4