Write a standalone calibration review bundle

Per-species review figures plus a one-page summary table, written to a plain directory that needs no Quarto render and no pipeline knowledge to inspect.

Usage

write_growth_review_bundle(
  dir,
  species,
  curves,
  candidate_curves = NULL,
  references,
  reference_curves = NULL,
  best,
  windows,
  scoring_file = "growth_scoring.csv",
  params_file = "the growth-parameter table",
  smooth_plots = TRUE,
  smooth_bin = 20L,
  smooth_site = NULL,
  x_max = NULL,
  density_min_plots = 500L,
  density_bins = 34L,
  density_points_max = 150L
)

Arguments

dir

Character. Output directory; created if absent.

species

Character vector of species to write.

curves

Simulated curves keyed by species (the verification run).

candidate_curves

Optional simulated curves for the best candidate.

references

Named list of reference tables, one per species.

reference_curves

Named list from growth_reference_curves(), one per species, supplying the age-binned ground-plot series.

best

A tibble of best-candidate parameters and fit statistics.

windows

A tibble from growth_fitting_windows().

scoring_file, params_file

Character. Paths named in the bundle's README.txt, so a reviewer is pointed at the files this project actually keeps them in.

smooth_plots

Logical. Overlay a spline through the ground-plot cloud on each panel, for comparison against the binned series. Display only; nothing is scored against it. See growth_smooth_observations().

smooth_bin, smooth_site

Passed to growth_smooth_observations().

x_max

Numeric or NULL. Upper age limit for every panel. NULL lets each panel extend to the last age in its own data.

density_min_plots

Integer. Draw the ground-plot cloud as a hex density rather than points once a species has at least this many admissible plots. Below it the hex grid is sparse and blocky and says less than the points did, so the switch is per species rather than a global default.

density_bins, density_points_max

Passed to plot_growth_candidate(): the number of hex bins across the panel, and the point cap used when the cloud is drawn as points instead. smooth_bin may be a single width or a vector named by species. Set smooth_site to the location column wherever the plots are a permanent network, so the fit and the binned series rest on the same evidence.

Value

Character vector of the written file paths.

See also

Other growth calibration helpers: extract_landscape_cohort_structures(), growth_add_objective(), growth_auto_window(), growth_best_candidates(), growth_bin_observations(), growth_calibration_design(), growth_calibration_partition(), growth_calibration_work_root(), growth_climatic_distance(), growth_climatic_weight(), growth_expand_over_pseudo_species(), growth_factorial_ratio_grid(), growth_fitting_windows(), growth_identifiability(), growth_inflation_factor(), growth_mortality_onset_frac(), growth_plot_palette(), growth_pseudo_species_name(), growth_reference_curves(), growth_reference_inflection(), growth_score_fit(), growth_scoring_for(), growth_smooth_observations(), growth_structure_cell_curves(), growth_structure_cohort_table(), growth_structure_design(), growth_structure_summary(), growth_window_for(), plot_growth_calibration(), plot_growth_candidate(), plot_growth_factorial_sensitivity(), plot_growth_structures(), read_growth_scoring(), read_landscape_cohort_structures(), scale_linetype_growth_reference()