Skip to contents

Main function to construct the TMB map argument for CEATTLE

Source: R/3-build_map.R
build_map.Rd

Builds the TMB map, which tells TMB which parameters to estimate and which to hold fixed (a fixed parameter is mapped to NA), and which parameters share a single estimated value. One helper handles each process block (recruitment, M1, predation, selectivity, catchability, ...).

Usage

build_map(
  data_list,
  params,
  debug = FALSE,
  random_rec = FALSE,
  random_sel = FALSE,
  random_q = NULL
)

Arguments

data_list

an Rceattle data_list

params

A parameter list created from build_params.

debug

Logical. If TRUE, fixes every parameter except the dummy (maps all to NA), so the model runs with no parameters estimated.

random_rec

Logical. If TRUE, treats recruitment deviations as random effects, meaning the variance parameter (R_log_sd) is estimated.

random_sel

Logical. If TRUE, treats selectivity deviations as random effects, meaning the variance parameter (sel_dev_log_sd) is estimated.

random_q

Logical. If TRUE, treats catchability deviations as random effects, meaning the variance parameter (index_q_dev_log_sd) is estimated. Defaults to data_list$random_q, which is what fit_mod() stores there.

Value

A list containing the factorized TMB map (mapFactor) and the original map matrix/array list (mapList).

Details

Fleets sharing a Selectivity_index or a Catchability_index have the lead fleet's map slice copied over the rest of the group, so a per-fleet setting that differs within a group is resolved here rather than reported. Those disagreements are checked in data_check(), which fit_mod() runs first; build_map() does not call it, so a caller invoking build_map() directly should run data_check() too.