Skip to contents

Turn climatic distance into a calibration weight

growth_climatic_weight.Rd

A weight rather than a cut-off, because any threshold is arbitrary and a plot just past it is not meaningfully different from one just inside. bandwidth is the distance at which a Gaussian weight falls to about 0.61; it is the one number to tune, and it can be chosen by cross-validation.

Usage

growth_climatic_weight(
  distance,
  bandwidth = 0.5,
  kernel = c("gaussian", "tricube", "uniform")
)

Arguments

distance

Numeric, from growth_climatic_distance().

bandwidth

Numeric. Scale over which similarity decays.

kernel

One of "gaussian" (smooth, never exactly zero), "tricube" (smooth, zero beyond bandwidth) or "uniform" (a hard cut-off at bandwidth).

Value

A numeric vector of weights in [0, 1].

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_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(), write_growth_review_bundle()