Predict Weight Over a Diameter Sequence by Grouping Factor

Summarise the fitted relationship between weight and diameter over a generated prediction grid: a sequence of diameter values crossed with the grouping factors named in by (one curve per group).

Usage

kb_predict_weight_by(
  fit,
  by = NULL,
  diameter = NULL,
  ...,
  new_levels = c("average", "sample"),
  conf_level = 0.95,
  estimate = stats::median,
  sig_fig = 3
)

Arguments

fit

A kb_fit_weight object.

by

A character vector of grouping factors, each drawn as a separate curve, or NULL for a single population-level curve. Each named factor is expanded over its observed levels and conditioned on its estimated random effects.

diameter

A numeric vector of sub-bulb diameters to predict over (in the same units as the fitted data), or NULL for an automatic sequence spanning the observed range.

...

These dots are for future extensions and must be empty.

new_levels

A string, one of "sample" or "average", controlling how random effects that are not conditioned on are treated (factors absent from the prediction, and any new level not seen in the fit). "sample" draws a new random effect from Normal(0, sd), widening the interval to include between-group variation; "average" holds the random effects at zero, giving the typical group. Known levels are always conditioned on. "sample" draws fresh randomness on each call, so set a seed with set.seed() for a reproducible interval.

conf_level

A number between 0 and 1 giving the compatibility-interval level.

estimate

A function that reduces a numeric vector of posterior draws to a scalar point estimate (e.g. median or mean).

sig_fig

A whole number of significant figures for summary output.

Value

A kb_predictions object: a summary tibble with estimate, lower, upper, the diameter predictor, and the by grouping columns.

Details

by selects the grouping factors that each get their own curve, conditioned on their estimated random effects. new_levels controls the factors not named in by. The default "average" holds those random effects at zero, giving the typical-group curve. "sample" instead draws a new random effect from its estimated distribution, widening the uncertainty to include between-group variation. Set a seed with set.seed() for reproducible CIs. The available by values are NULL (a single population curve), "site", and c("site", "year").

See also

kb_predict_weight() for predictions at the rows of a supplied data frame.

Other prediction: autoplot.kb_predictions(), kb_plot_predictions(), kb_predict_weight()

Examples

kb_predict_weight_by(fit_weight_sim_nereo, by = "site")
#> <kb_predictions> predictor: diameter | response: weight | by: site
#> # A tibble: 300 × 5
#>    site  diameter estimate  lower  upper
#>    <fct>    <dbl>    <dbl>  <dbl>  <dbl>
#>  1 site1     15.6   0.0163 0.0127 0.021 
#>  2 site1     18.1   0.0237 0.0194 0.0288
#>  3 site1     20.6   0.0327 0.0277 0.0387
#>  4 site1     23.2   0.0437 0.0376 0.0508
#>  5 site1     25.7   0.0567 0.0489 0.0652
#>  6 site1     28.2   0.0719 0.0626 0.0826
#>  7 site1     30.7   0.0897 0.0782 0.102 
#>  8 site1     33.2   0.11   0.0962 0.125 
#>  9 site1     35.7   0.133  0.116  0.151 
#> 10 site1     38.3   0.158  0.139  0.18  
#> # ℹ 290 more rows