Simulate temporal (multi-season) occupancy data

Colonization and extinction are constant across the n_seasons - 1 transition intervals by default. Supplying beta_gamma and/or beta_epsilon makes the corresponding rate SEASON-VARYING: a per-(site, interval) covariate is drawn into an [N x (n_seasons - 1)] matrix column (gamma_cov / eps_cov) of the returned data, and the rate is plogis(beta[1] + beta[2] * cov). Fit these with colonization = ~ gamma_cov / extinction = ~ eps_cov (the matrix column drives the interval-indexed rate).

Usage

simulate_dyn_occu(
  N = 100,
  J = 4,
  n_seasons = 5,
  beta_occ = c(0.5),
  beta_det = c(0),
  gamma = 0.2,
  epsilon = 0.1,
  beta_gamma = NULL,
  beta_epsilon = NULL,
  beta_det_season = NULL,
  seed = NULL
)

Arguments

N

Number of sites (default 100).

J

Number of visits per season (default 4).

n_seasons

Number of seasons (default 5).

beta_occ

Initial occupancy coefficients.

beta_det

Detection coefficients.

gamma

Colonization probability (default 0.2); ignored when beta_gamma is given.

epsilon

Extinction probability (default 0.1); ignored when beta_epsilon is given.

beta_gamma

Optional c(intercept, slope) for a season-varying colonization logit driven by a drawn per-(site, interval) covariate.

beta_epsilon

Optional c(intercept, slope) for a season-varying extinction logit driven by a drawn per-(site, interval) covariate.

beta_det_season

Optional c(intercept, slope) for a season-varying detection logit driven by a drawn per-(site, season) covariate det_cov (a [N x T] matrix column of data); fit with detection = ~ det_cov.

seed

Random seed.

Value

A list with y (3D array), data, and truth.

Examples

sim <- simulate_dyn_occu(N = 40, J = 3, n_seasons = 4, seed = 1)
dim(sim$y)