The function prepares the data, draws and alternative availabilites for estimation. All model components are placed in an estimation environment which provides the context for evaluating the log likelihood function.
prepare( ll, db, model_options, control = NULL, check_data = TRUE, debug = FALSE, ... )
A user supplied log-likelihood function
tibble() containing the data
A list of user supplied model options. This list is
validated within the function. See
validate for details.
A list of control options
A boolean equal to TRUE if we should run data checks and pad the data with NA.
A boolean equal to TRUE if you are debugging the log likelihood
function. It forces the number of cores to 1. Then you can simply
attach the estimation environment and the list of parameters
to run the log likelihood function line by line. The default is FALSE.
Additional named arguments that are added to the estimation environment. If cores > 1 the arguments are added to each individual estimation environment. No attempts at splitting across cores. This is useful if you need to pass e.g. a restriction matrix or other variables to the estimation environment so that they can be called by name/reference within the log-likelihood function.
The estimation environment or a list of estimation environments if cores > 1.