Make data for DRM stan models — make

This function creates the list used as the input for the stan model.

Usage

make_data(
  y,
  time,
  site,
  init_data = numeric(0),
  f_mort,
  m = -log(0.7),
  x_t,
  x_m,
  x_r,
  n_ages = 2,
  age_selectivity,
  ages_movement,
  adj_mat = matrix(0, ncol = 1, nrow = 1),
  .toggles,
  .priors,
  family = "gamma",
  reorder = TRUE,
  phi_hat = FALSE
)

Arguments

y

a numeric vector of species' densities.

time

a vector indicating the time point associated to each element of y.

site

a vector indicating the sites associated to each element of y.

init_data

an optional vector (of length n_ages - 1) to initialize the population dynamics.

f_mort

an optional matrix informing the instantaneous fishing mortality rates at each age (columns) and timepoint (rows).

m

a numeric value corresponding to the instantaneous natural mortality rate. The default value for this is -log(.7), as it implies a survival rate of 0.70 between age classes.

x_t

a design matrix of variables associated to the probability of absence at each site/time.

x_m

a design matrix of variables associated to survival.

x_r

a design matrix of variables associated to recruitment.

n_ages

an integer indicating the number of ages for the underlying population dynamic model.

age_selectivity

a numeric vector with n_ages elements, where each element indicates the selectivity of a respective age. All the elements of this vector must lie between 0 and 1.

ages_movement

An integer or a numeric vector specifying the ages at which individuals of the focal species are assumed to move. If ages_movement is an integer, individuals younger than this age are considered static (non-moving). If ages_movement is a numeric vector of length n_ages, it indicates movement capability for each age group. A value of 0 indicates the corresponding age group is static, while 1 indicates movement is allowed. For example, c(0, 0, 1, 1, 0) specifies that age groups 1, 2, and 5 are static, while 3 and 4 are mobile.

adj_mat

an adjacency matrix of dimensions sites \(\times\) sites. Its elements are 1 if two sites are neighbors and zero otherwise.

.toggles

a list of toggles for model components. The components are:

rho_mu: 1 to explicitly relate rho to mu and 0 otherwise.
cloglog: 1 to use the complementary log-log and 0 for the logit link function for the absence probabilities.
movement: 1 to allow for (adjacent) movement; 0 for static.
est_surv: 1 to estimate mortality and 0 otherwise.
est_init: 1 to estimate initial values for lambda and 0 otherwise.
minit: 1 to use mortality to estimate initial age classes and 0 otherwise.
ar_re: a character. It assumes one of the following values: "none" - no AR, "rec" AR(1) for recruitment, "surv" - AR(1) for survival (only works when est_surv is on), "dens" - AR(1) for density.
iid_re: a character. It assumes one of the following values: "none" - no iid re, "rec" iid re for recruitment, "surv" - iir re for survival (only works when est_surv is on), "dens" - iid_re for density.
sp_re: a character. It assumes one of the following values: "none" - no ICAR re, "rec" ICAR re for recruitment, "surv" - ICAR re for survival (only works when est_surv is on), "dens" - ICAR_re for density.

.priors

a list of priors hyperparameters.

family

a character specifying the family of the probability distribution assumed for density. The options are:

"gamma" (default): gamma parametrized in terms of its mean;
"lognormal": log-normal parametrized in terms of its mean;
"loglogistic": log-logistic parametrized in terms of its median (usual parametrization);
"lognormal_legacy": log-normal with its usual parametrization;

reorder

a boolean telling whether the data needs to be reordered. The default is TRUE and means the data points will be ordered by site and time, respectively.

phi_hat

a boolean indicating whether the prior on phi should be determined through the data (using get_phi_hat()).

Value

A list containing the data and settings to be used as input for the Stan model.

Author

lcgodoy