Fit an xsdm model

Usage

fit_xsdm(
  xsdm_object,
  fit = NULL,
  recompile = FALSE,
  normalize = FALSE,
  prior_parameters = list(mu_par_1 = NULL, mu_par_2 = NULL, sigl_par = NULL, sigr_par =
    NULL, c_par_1 = NULL, c_par_2 = NULL, L_par = NULL),
  nchains = NULL,
  compile_standalone = FALSE,
  ...
)

Arguments

xsdm_object: an xsdm object created with constuctor and validated
fit: String. Default is TRUE. Calculate parameters either by sampling, optimization or Laplace methods. Calculate the mle estimation (default) or Maximum a posteriori estimation (need to pass jacobian = TRUE argument). Possible parameters are c("mle", "map", "mle.laplace", "map.laplace")
recompile: Default FALSE. If TRUE, recompile the Lewontin-Cohen model to use features not availiable out of the box. It is needed to use laplace and optim methods.
normalize: Logical, normalize the time series using the mean and the standard deviation of the presence-absence provided points
prior_parameters: list with parameters passed to prior.
nchains: Default NULL. The number of mcmc chains passed to cmdstanr sample function.
compile_standalone: Default TRUE. It is useful to export loglik functions
...: argument passed to stan model sampling

Value

a fitted xsdm object