Set up survey parameterization for the operating model simulation

Sets the survey catchability, selectivity, timing, index type and the age and length composition settings, with their overdispersion and correlation parameters. Call after Setup_Sim_Dim.

Usage

Setup_Sim_Survey(
  sim_list,
  srv_sel_input,
  ObsSrvIdx_SE = array(0.2, dim = c(sim_list$n_regions, sim_list$n_yrs, sim_list$n_seas,
    sim_list$n_srv_fleets)),
  ln_sigmaSrvIdxAA = array(log(0.2), dim = c(sim_list$n_obs_ages, sim_list$n_sexes,
    sim_list$n_srv_fleets)),
  UseSrvIdxAA = array(0, dim = c(sim_list$n_regions, sim_list$n_yrs, sim_list$n_seas,
    sim_list$n_obs_ages, sim_list$n_sexes, sim_list$n_srv_fleets)),
  ObsSrvIdxAA_SE = NULL,
  SrvIdxAA_Type = "spltRaggS",
  SrvIdxAA_LikeType = "lognormal",
  SrvIdxAA_sigma_form = "none",
  use_srv_idx_aa = rep(0, sim_list$n_srv_fleets),
  AgeObsCorr_srv_idx = "iid",
  trans_rho_srv_idx = NULL,
  trans_rho_srv_idx_year = NULL,
  trans_rho_srv_idx_us = NULL,
  SrvIdxAA_seas_Type = 0,
  UseSrvIdxAA_pop = NULL,
  ln_sigmaSrvIdxAA_pop = NULL,
  ObsSrvIdxAA_pop_SE = NULL,
  SrvIdxAA_pop_Type = "spltRaggS",
  SrvIdxAA_pop_LikeType = "lognormal",
  SrvIdxAA_pop_sigma_form = "none",
  SrvIdxAA_pop_seas_Type = 0,
  AgeObsCorr_srv_idx_pop = "iid",
  trans_rho_srv_idx_pop = NULL,
  trans_rho_srv_idx_pop_year = NULL,
  trans_rho_srv_idx_pop_us = NULL,
  ObsSrvIdx_pop_SE = array(0.2, dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_yrs, sim_list$n_seas, sim_list$n_srv_fleets)),
  srv_q_input = array(1, dim = c(sim_list$n_regions, sim_list$n_yrs,
    sim_list$n_srv_fleets, sim_list$n_sims)),
  t_srv = array(1, dim = c(sim_list$n_regions, sim_list$n_seas, sim_list$n_srv_fleets)),
  srv_idx_type = array(1, dim = c(sim_list$n_srv_fleets)),
  srv_idx_ages = NULL,
  SrvIdx_LikeType = rep(0, sim_list$n_srv_fleets),
  sigmaSrvIdx_form = "fix",
  ln_sigmaSrvIdx = NULL,
  sigmaSrvIdx_pop_form = "fix",
  ln_sigmaSrvIdx_pop = NULL,
  SrvIdx_seas_Type = NULL,
  SrvIdx_pop_seas_Type = NULL,
  SrvAgeComps_seas_Type = NULL,
  SrvLenComps_seas_Type = NULL,
  SrvAgeComps_pop_seas_Type = NULL,
  SrvLenComps_pop_seas_Type = NULL,
  seas_agg_slot = NULL,
  SrvIdx_Cov = NULL,
  UseSrvIdx = NULL,
  comp_srv_caal_like = rep(999, sim_list$n_srv_fleets),
  ISS_Srv_caal = NULL,
  ln_Srv_caal_theta = NULL,
  ln_Srv_caal_theta_agg = NULL,
  Srv_caal_Type = array(999, dim = c(sim_list$n_yrs, sim_list$n_srv_fleets)),
  comp_srvage_like = rep(0, sim_list$n_srv_fleets),
  ISS_SrvAgeComps = array(100, dim = c(sim_list$n_regions, sim_list$n_yrs,
    sim_list$n_seas, sim_list$n_sexes, sim_list$n_srv_fleets, sim_list$n_sims)),
  ln_SrvAge_theta = array(log(1), dim = c(sim_list$n_regions, sim_list$n_sexes,
    sim_list$n_srv_fleets)),
  ln_SrvAge_theta_agg = rep(log(1), sim_list$n_srv_fleets),
  SrvAge_corr_pars_agg = rep(0.01, sim_list$n_srv_fleets),
  SrvAge_corr_pars = array(0.01, dim = c(sim_list$n_regions, sim_list$n_sexes,
    sim_list$n_srv_fleets, 2)),
  SrvAgeComps_Type = array(2, dim = c(sim_list$n_yrs, sim_list$n_srv_fleets)),
  comp_srvlen_like = rep(0, sim_list$n_srv_fleets),
  ISS_SrvLenComps = array(100, dim = c(sim_list$n_regions, sim_list$n_yrs,
    sim_list$n_seas, sim_list$n_sexes, sim_list$n_srv_fleets, sim_list$n_sims)),
  ln_SrvLen_theta = array(log(1), dim = c(sim_list$n_regions, sim_list$n_sexes,
    sim_list$n_srv_fleets)),
  ln_SrvLen_theta_agg = rep(log(1), sim_list$n_srv_fleets),
  SrvLen_corr_pars_agg = rep(0.01, sim_list$n_srv_fleets),
  SrvLen_corr_pars = array(0.01, dim = c(sim_list$n_regions, sim_list$n_sexes,
    sim_list$n_srv_fleets, 2)),
  SrvLenComps_Type = array(2, dim = c(sim_list$n_yrs, sim_list$n_srv_fleets)),
  comp_srvage_pop_like = rep(0, sim_list$n_srv_fleets),
  ISS_SrvAgeComps_pop = array(100, dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_yrs, sim_list$n_seas, sim_list$n_sexes, sim_list$n_srv_fleets,
    sim_list$n_sims)),
  ln_SrvAge_pop_theta = array(log(1), dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_sexes, sim_list$n_srv_fleets)),
  ln_SrvAge_pop_theta_agg = array(log(1), dim = c(sim_list$n_pop, sim_list$n_srv_fleets)),
  SrvAge_pop_corr_pars = array(0.01, dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_sexes, sim_list$n_srv_fleets, 2)),
  SrvAge_pop_corr_pars_agg = array(0.01, dim = c(sim_list$n_pop, sim_list$n_srv_fleets)),
  SrvAgeComps_pop_Type = array(2, dim = c(sim_list$n_yrs, sim_list$n_srv_fleets)),
  comp_srvlen_pop_like = rep(0, sim_list$n_srv_fleets),
  ISS_SrvLenComps_pop = array(100, dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_yrs, sim_list$n_seas, sim_list$n_sexes, sim_list$n_srv_fleets,
    sim_list$n_sims)),
  ln_SrvLen_pop_theta = array(log(1), dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_sexes, sim_list$n_srv_fleets)),
  ln_SrvLen_pop_theta_agg = array(log(1), dim = c(sim_list$n_pop, sim_list$n_srv_fleets)),
  SrvLen_pop_corr_pars = array(0.01, dim = c(sim_list$n_pop, sim_list$n_regions,
    sim_list$n_sexes, sim_list$n_srv_fleets, 2)),
  SrvLen_pop_corr_pars_agg = array(0.01, dim = c(sim_list$n_pop, sim_list$n_srv_fleets)),
  SrvLenComps_pop_Type = array(2, dim = c(sim_list$n_yrs, sim_list$n_srv_fleets)),
  SrvLenComps_sel = rep("age", sim_list$n_srv_fleets),
  srv_sel_l_input = NULL
)

Arguments

sim_list: Simulation list returned by Setup_Sim_Dim.
srv_sel_input: Survey selectivity array [n_pop x n_regions x n_yrs x n_seas × n_ages × n_sexes × n_srv_fleets × n_sims]. No default.
ObsSrvIdx_SE, ObsSrvIdx_pop_SE: Lognormal observation error sd for the survey indices, [n_regions × n_yrs × n_seas × n_srv_fleets] with a leading n_pop for the second. Default 0.2.
ln_sigmaSrvIdxAA: Log-scale observation error for the index at age, `n_obs_ages x n_sexes x n_srv_fleets`. The sex dim is required.
UseSrvIdxAA: Integer array `n_regions x n_yrs x n_seas x n_obs_ages x n_sexes x n_srv_fleets`, `1` where a survey index at age is drawn, on the observed ages `AgeingError_srv_input` reads onto. The sex dim is required: a data source summed over sexes has its flag in sex slot one.
ObsSrvIdxAA_SE: Reported standard errors shaped like `UseSrvIdxAA`, read only when `SrvIdxAA_sigma_form` asks for them.
SrvIdxAA_Type: Which dims each fleet reports separately: `"agg"`, `"spltRaggS"` (default), `"aggRspltS"` or `"spltRspltS"`.
SrvIdxAA_LikeType: `"lognormal"` (default) or `"normal"`, per fleet.
SrvIdxAA_sigma_form: Where the observation error comes from: `"none"` (default), `"data"`, `"est_additive"` or `"est_quadrature"`.
use_srv_idx_aa: Integer vector `n_srv_fleets`, `1` for fleets whose index at age is drawn.
AgeObsCorr_srv_idx: How each fleet's index-at-age residuals are correlated, as in [Setup_Mod_SrvIdx_and_Comps()]: `"iid"` (default), `"1dar1"`, `"us"` or `"2dar1"`; see [Setup_Sim_Fishing()].
trans_rho_srv_idx, trans_rho_srv_idx_year, trans_rho_srv_idx_us: Unconstrained correlations under the estimation model's names and shapes, as for the fishery's. `NULL` (default) is zero.
SrvIdxAA_seas_Type, SrvIdxAA_pop_seas_Type: Per fleet, `1` where the index at age is a year total, drawn once a year from every season summed; see [Setup_Sim_Fishing()]. `0` (default) draws each season.
UseSrvIdxAA_pop, ln_sigmaSrvIdxAA_pop, ObsSrvIdxAA_pop_SE, SrvIdxAA_pop_Type, SrvIdxAA_pop_LikeType, SrvIdxAA_pop_sigma_form, AgeObsCorr_srv_idx_pop, trans_rho_srv_idx_pop, trans_rho_srv_idx_pop_year, trans_rho_srv_idx_pop_us: The population-specific index at age, as the aggregated one with a leading `n_pop` dim on the arrays and parameters, each population drawn on its own. `NULL` draws none.
srv_q_input: Survey catchability array [n_regions × n_yrs × n_srv_fleets × n_sims]. Default 1.
t_srv: Survey timing as a fraction of the year or season, [n_regions × n_seas × n_srv_fleets]. Default 1.
srv_idx_type: Index type per fleet: 0/"abd" or 1/"biom" (default).
srv_idx_ages: Ages counted in each fleet's index total, a 0/1 array [n_ages × n_srv_fleets], the estimation model's srv_idx_ages. NULL (default) counts every age.
SrvIdx_LikeType: Error structure each fleet's index is drawn under: "lognormal" (0, default), "normal" (1) or "mvn" (2), matching the estimation model. An mvn fleet draws from SrvIdx_Cov through a common-factor decomposition (see cov_to_factor) instead of ObsSrvIdx_SE, and its population-specific data source stays lognormal.
sigmaSrvIdx_form, sigmaSrvIdx_pop_form: How the index sd combines the reported errors with an estimated part, as sigmaSrvIdx_spec in Setup_Mod_SrvIdx_and_Comps: "fix" (0, default), "est_additive" (1), "est_quadrature" (2) or "est_replace" (3). See Setup_Sim_Fishing.
ln_sigmaSrvIdx, ln_sigmaSrvIdx_pop: Log of the estimated part, one per survey, read under a form other than "fix". Default NULL.
SrvIdx_seas_Type, SrvIdx_pop_seas_Type, SrvAgeComps_seas_Type, SrvLenComps_seas_Type, SrvAgeComps_pop_seas_Type, SrvLenComps_pop_seas_Type: Whether the operating model reports a survey data source once a season ("spltSeas", the default) or once a year as a season total ("aggSeas"), one value for every survey or one per survey. An annual total is written into the season seas_agg_slot names, season one by default, with the other seasons left at zero and the observation error applied once to that total. A composition is drawn from the numbers summed over the year. An estimation model reading it should mark the same season in its Use array and set the matching argument in Setup_Mod_SrvIdx_and_Comps.
seas_agg_slot: Named list, by survey data source ("SrvIdx", "SrvAgeComps_pop" and so on), of the season each survey's year total is written into, an integer matrix [n_yrs x n_srv_fleets]. See Setup_Sim_Fishing. Default NULL.
SrvIdx_Cov: List with one element per fleet holding the fixed covariance over that fleet's fitted index observations, ordered by scanning UseSrvIdx in array order. Required for mvn fleets. Default NULL.
UseSrvIdx: Fit flags [n_regions x n_yrs x n_seas x n_srv_fleets] from the estimation model, used to place each simulated cell in the covariance. Its year dim may be shorter than the simulation, in which case later years draw with the mean factor scale and loading. Required for mvn fleets. Default NULL.
comp_srv_caal_like: Conditional age-at-length likelihood per fleet: `"Multinomial"` (0), `"Dirichlet-Multinomial"` (1) or `"none"` (999, default). The survey twin of `comp_fish_caal_like`, and only these two families exist for CAAL.
ISS_Srv_caal: Number of fish aged within each length row, `n_regions x n_yrs x n_seas x n_caal_lens x n_sexes x n_srv_fleets x n_sims`. A bin whose sample size rounds to zero is skipped. `NULL` (default) draws no CAAL; supplying it alongside a likelihood other than `"none"` switches `do_srv_caal` on. Requires `n_lens`.
ln_Srv_caal_theta: Log overdispersion for the Dirichlet-multinomial, `n_regions x n_sexes x n_srv_fleets`, read under the split types and ignored under the multinomial. Default log(1).
ln_Srv_caal_theta_agg: The aggregated type's counterpart, length `n_srv_fleets`. Default log(1).
Srv_caal_Type: Composition structure per year and fleet, `n_yrs x n_srv_fleets`, with the codes of `Fish_caal_Type`: `"agg"` (0), `"spltRspltS"` (1), `"spltRjntS"` (2) or `"none"` (999, default). The simulator takes the year by fleet array directly.
comp_srvage_like, comp_srvlen_like, comp_srvage_pop_like, comp_srvlen_pop_like: Composition likelihood per fleet for the four survey composition data sources: 0/"Multinomial" (default), 1/"Dirichlet-Multinomial", 2/"iid-Logistic-Normal", 3/"1d-Logistic-Normal" or 4/"2d-Logistic-Normal".
ISS_SrvAgeComps, ISS_SrvLenComps: Input sample sizes [n_regions × n_yrs × n_seas × n_sexes × n_srv_fleets × n_sims]. Default 100.
ln_SrvAge_theta, ln_SrvLen_theta: Log-scale overdispersion [n_regions × n_sexes × n_srv_fleets], read under likelihoods 1-4. Default log(1).
ln_SrvAge_theta_agg, ln_SrvLen_theta_agg: The aggregated types' counterparts, length n_srv_fleets. Default log(1).
SrvAge_corr_pars_agg, SrvLen_corr_pars_agg: The aggregated types' counterparts, length n_srv_fleets, read under likelihood 3. Default 0.01.
SrvAge_corr_pars, SrvLen_corr_pars: Correlation parameters [n_regions × n_sexes × n_srv_fleets × 2], the age AR1 and the sex correlation, read under likelihoods 3 and 4. Default 0.01.
SrvAgeComps_Type, SrvLenComps_Type, SrvAgeComps_pop_Type, SrvLenComps_pop_Type: Composition structure [n_yrs × n_srv_fleets]: 0/"agg", 1/"spltRspltS", 2/"spltRjntS" (default) or 999/"none".
ISS_SrvAgeComps_pop, ISS_SrvLenComps_pop: The population-specific counterparts, with a leading n_pop dim. Default 100.
ln_SrvAge_pop_theta, ln_SrvLen_pop_theta: Log-scale overdispersion for the population-specific data sources [n_pop × n_regions × n_sexes × n_srv_fleets]. Default log(1).
ln_SrvAge_pop_theta_agg, ln_SrvLen_pop_theta_agg: Their aggregated counterparts [n_pop × n_srv_fleets]. Default log(1).
SrvAge_pop_corr_pars, SrvLen_pop_corr_pars: Correlation parameters for the population-specific data sources [n_pop × n_regions × n_sexes × n_srv_fleets × 2]. Default 0.01.
SrvAge_pop_corr_pars_agg, SrvLen_pop_corr_pars_agg: Their aggregated counterparts [n_pop × n_srv_fleets]. Default 0.01.
SrvLenComps_sel: Character vector [n_srv_fleets], "age" (default) or "length", as in Setup_Mod_SrvIdx_and_Comps. "length" spreads the numbers present at each age over length and selects them length by length, so the length compositions read srv_sel_l_input.
srv_sel_l_input: Survey selectivity at length [n_regions × n_yrs × n_lens × n_sexes × n_srv_fleets × n_sims], read under SrvLenComps_sel = "length".

Value

sim_list with the survey fields appended: $srv_sel, $srv_q, $ObsSrvIdx_SE, $ObsSrvIdx_pop_SE, $t_srv, $srv_idx_type, and, for each of the four composition data sources, its likelihood, input sample sizes, overdispersion, correlation parameters and composition type. Character codes are converted to integers before storage.

Set up survey parameterization for the operating model simulation

Usage

Arguments

Value

See also