Simulate an open-population distance-sampling data set — simulate_distsamp

Draws from the distsamp_open() model: an open metapopulation (N_1 ~ Poisson / NB(lambda), N_t = Binomial(N_{t-1}, omega) + Poisson(gamma)) observed by distance sampling (half-normal key, scale sigma) at each primary period.

Usage

simulate_distsamp_open(
  N = 200,
  cutpoints = c(0, 10, 20, 30, 40),
  n_seasons = 4L,
  transect = "line",
  n_abund_covs = 1,
  n_det_covs = 1,
  beta_lambda = NULL,
  beta_sigma = NULL,
  omega = 0.7,
  gamma = 2.5,
  mixture = c("poisson", "negbin", "zip", "zinb"),
  size = 3,
  zi = 0.3,
  dynamics = c("constant", "notrend", "trend", "autoreg", "ricker", "gompertz"),
  K = NULL,
  r = 0.3,
  seed = NULL
)

Arguments

N: Number of sites (default 200).
cutpoints: Distance-bin edges 0 = c_0 < ... < c_B.
n_seasons: Number of primary periods (default 4).
transect: "line" (default) or "point".
n_abund_covs, n_det_covs: Number of abundance / distance covariates.
beta_lambda, beta_sigma: Coefficients on the log-abundance and log-scale arms. Defaults give moderate abundance / detection.
omega, gamma: Apparent survival probability and recruitment rate (intercept-only defaults 0.7 / 2.5).
mixture: Initial-abundance distribution: "poisson" (default), "negbin" / "zinb" (draws N_1 negative-binomial with size size), or "zip" / "zinb" (a zi fraction of sites are structural zeros).
size: Negative-binomial size (overdispersion r) when mixture draws an NB initial abundance (default 3).
zi: Structural-zero probability when mixture is "zip" / "zinb" (default 0.3).
dynamics: Population-dynamics form (see distsamp_open()): "constant" (default), "notrend", "trend", "autoreg", "ricker", or "gompertz". The alternative dynamics draw the abundance sequence from the matching transition (Poisson initial abundance only).
K: Carrying capacity for "ricker" / "gompertz" (default 4 * lambda0); r the intrinsic growth rate for those forms (default 0.3). For "trend" / "autoreg" the recruitment multiplier is gamma.
r: Intrinsic growth rate for "ricker" / "gompertz" (default 0.3).
seed: Optional random seed.

Value

A list with y ([n_sites x n_bins x n_seasons] distance-band counts), data, and truth.

Examples

sim <- simulate_distsamp_open(N = 40, n_seasons = 3, seed = 1)
dim(sim$y)