Skip to contents

Sets the retention selectivity forms, time blocks, continuous time variation, process error, annual deviations and fixed or estimated parameters. Time variation through cont_tv_ret_sel and blocks through ret_sel_blocks are mutually exclusive within a fleet. Call after Setup_Mod_FishIdx_and_Comps.

Usage

Setup_Mod_Retsel(
  input_list,
  cont_tv_ret_sel,
  ret_sel_blocks,
  ret_sel_model,
  retsel_pe_pars_spec,
  ret_fixed_sel_pars_spec,
  ret_sel_devs_spec,
  ret_sel_corr_opt_semipar,
  Use_ret_selex_prior,
  ret_selex_prior,
  retsel_devs_shared_bins,
  ret_selex_type,
  use_fixed_ret_sel,
  ret_sel_input,
  ret_sel_bin_dev_bins = NULL,
  cont_tv_retsel_bin_devs = rep("none", input_list$data$n_fish_fleets),
  retsel_pe_wt = rep(1, input_list$data$n_fish_fleets),
  retsel_rw_init_sigma = rep(5, input_list$data$n_fish_fleets),
  retsel_dont_est_dev_first = rep(0, input_list$data$n_fish_fleets),
  ret_sel_nonpar_est_bins,
  ret_sel_sex_offset = rep("none", input_list$data$n_fish_fleets),
  ...
)

Arguments

input_list

Named list with $data, $par, $map and $verbose.

cont_tv_ret_sel

Character vector of continuous time variation per fleet, each "<type>_Fleet_<f>": "none", "iid", "rw", "3dmarg", "3dcond" or "2dar1". Stored as integer codes.

ret_sel_blocks

Character vector of discrete blocks, either "none_Fleet_<f>" or "Block_<b>_Year_<start>-<end>_Fleet_<f>" with "terminal" allowed as the end year. Expanded into an [n_regions × n_years × n_fish_fleets] array.

ret_sel_model

Character vector of the retention form per fleet, and optionally per block: "<type>_Fleet_<f>" or "<type>_Fleet_<f>_Block_<b>". The forms and their syntax are those of fish_sel_model in Setup_Mod_Fishsel_and_Q.

retsel_pe_pars_spec

Process error parameters for time-varying retention, one entry per fishery fleet.

ret_fixed_sel_pars_spec

Which fixed retention parameters are estimated.

ret_sel_devs_spec

Structure of the annual retention deviations per fleet.

ret_sel_corr_opt_semipar

Optional correlation structure for semi-parametric retention deviations, one entry per fishery fleet.

Use_ret_selex_prior

Integer flag (0/1) for retention selectivity priors.

ret_selex_prior

Data frame with columns region, fleet, block, sex, par, mu, sd and an optional type ("par" or "value"), as fish_selex_prior in Setup_Mod_Fishsel_and_Q.

retsel_devs_shared_bins

Bins sharing one retention deviation series.

ret_selex_type

Character scalar, "age" or "length".

use_fixed_ret_sel

Binary array [n_pop × n_regions × n_years × n_seas × n_fish_fleets], 1 to use ret_sel_input.

ret_sel_input

Fixed retention array [n_pop × n_regions × n_years × n_seas × n_bins × n_sexes × n_fish_fleets].

ret_sel_bin_dev_bins

List with one element per fleet naming the bins that fleet overrides, or NULL for none. An overridden bin takes a freely estimated annual value in place of what the functional form produced, applied after every other transformation including standardization. Default NULL.

cont_tv_retsel_bin_devs

Character vector [n_fish_fleets] of the process error on the bin-override deviations: "none" (default), "iid" or "rw".

retsel_pe_wt

Numeric vector [n_fish_fleets] multiplying the retention process error likelihood. Default 1. 0 skips that fleet's process error, so the deviations stay estimated but enter the objective only through the data and any smoothness or centering penalties. Values other than 0 or 1 make an estimated process error sigma reinterpretable. Applies to ln_retsel_devs only.

retsel_rw_init_sigma

Numeric vector [n_fish_fleets] giving the standard deviation of the first year of an "rw" deviation series. Default 5, which leaves that year effectively free. NA instead starts the walk at zero under the walk's own estimated sigma.

ret_sel_nonpar_est_bins

Estimated bins for non-parametric retention.

ret_sel_sex_offset

Character vector [n_fish_fleets] linking the sexes of a fleet's retention curve: "none" (default), "par", "scale" or "par_scale", as fish_sel_sex_offset in Setup_Mod_Fishsel_and_Q. Retention is a fraction, so a scale offset only makes sense where the scaled curve stays at or below one, and nothing enforces that.

...

Optional starting values for the selectivity parameters and deviations.

Value

input_list with the parsed retention structure arrays in $data, the starting values in $par and the factor maps in $map.