Haines 2019 provides likelihood (density) calculation and data generating (randomization) functions in R for multinomial N-mixture models. Here, I translate those models into nimble code for use alongside the binomial N-mixture models available in nimbleEcology.
Haines' work extends the fast-algorithm approach that nimbleEcology uses with binomial N-mixture models by use of recursion to avoid infinite sums.