Tidy long-format derived quantities from an Rceattle fit

Returns the model's derived population quantities in long form so that custom plots and post-processing don't have to walk the deeply nested quantities list or inherit the dimnames decisions in rename_output(). Two shapes are flattened into one tidy frame: species-by-year quantities (e.g. biomass, ssb, R, F_spp) and species-by-sex-by-age-by-year quantities (e.g. N_at_age, biomass_at_age, M_at_age). For the species-level shapes, sex and age are returned as NA. Cells of the 4D arrays that are padded out to max(nsex) / max(nages) for species with fewer sexes or ages are dropped.

Usage

# S3 method for class 'Rceattle'
as.data.frame(
  x,
  row.names = NULL,
  optional = FALSE,
  which = c("biomass", "ssb", "R", "biomass_depletion", "ssb_depletion", "F_spp"),
  ci_level = 0.95,
  ...
)

Arguments

x

An object of class "Rceattle" returned by fit_mod().

row.names, optional

Ignored; present for the as.data.frame() generic.

which

Character vector of quantity names to extract, or "all" for every quantity with a known shape. Defaults to a common population-level summary. See names(Rceattle:::.RCEATTLE_QUANTITIES) for the full list.

ci_level

Confidence level for lwr and upr. Default 0.95.

...

Currently unused.

Value

A data.frame with columns year, species, sex, age, quantity, value, se, lwr, upr. species is the character species name from data_list$spnames. Rows are sorted in the order which was given.

Details

Standard errors (se) and confidence intervals (lwr, upr) are populated from the TMB sdreport for any quantity that was ADREPORT'd (biomass, ssb, R, exploitable_biomass, biomass_depletion, ssb_depletion); other quantities and fits produced with getsd = FALSE get NA for se / lwr / upr. Set ci_level to widen or narrow the band.

A strictly positive series takes its interval on the log scale, so it is right-skewed and cannot reach zero. This is the same construction plot_timeseries() draws, so the table and the figure agree.