Simulate a time-to-detection occupancy data set — simulate_occu

Draws from the occu_ttd() model: site occupancy z ~ Bernoulli(psi) and, at occupied sites, exponential time-to-detection at rate lambda over a survey of length Tmax (a non-detection is censored at Tmax).

Usage

simulate_occu_ttd(
  N = 200,
  J = 4,
  n_psi_covs = 1,
  n_rate_covs = 1,
  beta_psi = NULL,
  beta_rate = NULL,
  Tmax = 3,
  seed = NULL
)

Arguments

N: Number of sites (default 200).
J: Number of replicate surveys per site (default 4).
n_psi_covs, n_rate_covs: Number of occupancy / rate covariates.
beta_psi: Occupancy coefficients c(intercept, slopes...) (logit). Default c(qlogis(0.6), runif(n_psi_covs, -0.5, 0.5)).
beta_rate: Log detection-rate coefficients c(intercept, slopes...). Default c(log(0.6), runif(n_rate_covs, -0.4, 0.4)).
Tmax: Survey length (scalar). Default 3.
seed: Optional random seed.

Value

A list with y (N x J TTD matrix), data, Tmax, and truth.

Examples

sim <- simulate_occu_ttd(N = 50, J = 3, seed = 1)
dim(sim$y)