Skip to contents

Run Iterative Francis Reweighting Procedure

run_francis.Rd

Runs an iterative Francis reweighting procedure for pooled and population-specific fishery and survey age and length compositions. Repeatedly fits the model and updates composition weights until the specified number of iterations is reached.

Usage

run_francis(
  data,
  parameters,
  mapping,
  random = NULL,
  n_francis_iter = 10,
  newton_loops = 0
)

Arguments

data

A list of model input data containing observed compositions, usage flags, input sample sizes, and weight arrays for both pooled (Wt_FishAgeComps, Wt_FishLenComps, Wt_SrvAgeComps, Wt_SrvLenComps, Wt_FishAgeComps_discard, Wt_FishLenComps_discard) and population-specific (Wt_FishAgeComps_pop, Wt_FishLenComps_pop, Wt_SrvAgeComps_pop, Wt_SrvLenComps_pop, Wt_FishAgeComps_discard_pop, Wt_FishLenComps_discard_pop) data streams.

parameters

A list of model parameters passed to fit_model.

mapping

A list or mapping object passed to fit_model.

random

Character vector of random effects passed to fit_model. Default NULL.

n_francis_iter

Integer. Number of Francis reweighting iterations. Default 10.

newton_loops

Integer. Number of Newton refinement steps passed to fit_model. Default 0.

Value

A named list with:

obj

The fitted model object from the final iteration, augmented with $data, $parameters, $mapping, $random, and $rep.

start_mean_francis

Mean composition fits from the first iteration, as returned by do_francis_reweighting.

end_mean_francis

Mean composition fits from the final iteration.

recorded_weights

Long-format dataframe of Francis weights from every iteration for all pooled and population-specific data streams, with columns Region, Year, Seas, Sex, Fleet, Weight, Type, Pop (pooled rows have Pop = NA), and iter.

See also

Other Francis Reweighting: do_francis_reweighting()

Examples

if (FALSE) { # \dontrun{
  out <- run_francis(data = data, parameters = parameters,
                     mapping = mapping, random = NULL,
                     n_francis_iter = 5, newton_loops = 3)
  out$obj
  out$end_mean_francis
  out$recorded_weights
} # }