Utility for summarising uncertainty — do_uncertainty

This function is not meant to be used by typical users. It can be used to build custom pipes with estimateR.

do_uncertainty_summary(
  original_values,
  bootstrapped_values,
  uncertainty_summary_method,
  value_col,
  bootstrap_id_col,
  index_col,
  output_Re_only,
  combine_bootstrap_and_estimation_uncertainties = FALSE,
  Re_HPDs = NULL,
  ...
)

Arguments

original_values	Optional. Values of reference used to construct the uncertainty interval around. Typically, these are estimates obtained on the original data. Must be a dataframe with a timestep index column named `index_col` and a value column named `value_col`. The index column must not contain any `NA` value.
bootstrapped_values	Bootstrap replicates of the original data. Must be a dataframe in the long format with a timestep index column named `index_col`, a bootstrap replicate index column named `bootstrap_id_col`, and a value column named `value_col`. The index column must not contain any `NA` value.
uncertainty_summary_method	One of these options: 'original estimate - CI from bootstrap estimates'. The confidence interval is built using bootstrapped estimates and centered around the original estimates. 'bagged mean - CI from bootstrap estimates'. The confidence interval is built using bootstrapped estimates and centered around the mean of bootstrapped estimates and original estimates.
value_col	string. Name of the column containing values.
bootstrap_id_col	string. Name of the column containing bootstrap samples numbering. Id 0 must correspond to values associated to the original data.
index_col	string. Name of the index column. The index tracks which data point in bootstrapped values corresponds to which data point in the original values.
output_Re_only	boolean. Should the output only contain Re estimates? (as opposed to containing results for each intermediate step)
combine_bootstrap_and_estimation_uncertainties	boolean. Combine uncertainty from Re estimation with uncertainty from observation process? If `TRUE`, the credible intervals for Re estimates must be passed via `Re_HPDs`. The output credible intervals will be the union of bootstrapping intervals and Re estimation intervals.
Re_HPDs	Optional. Credible intervals for Re estimates. Use only if `combine_bootstrap_and_estimation_uncertainties` is `TRUE`.
...	Arguments passed on to `.summarise_CI_bootstrap` `central_values` Values around which the confidence interval is going to be centered. Must be a dataframe with a timestep index column named `index_col` and a value column named `value_col`. The index column must not contain any `NA` value. `alpha` value between 0 and 1. Confidence level of the confidence interval. `prefix_up` string. prefix to add to `output_value_col` to name the column containing the upper limit of confidence interval `prefix_down` string. prefix to add to `output_value_col` to name the column containing the lower limit of confidence interval

Value

A dataframe containing Re estimates (column 'Re_estimate') and confidence interval boundaries, with 4 columns like so:

index_col, the timestep index column
A column named output_value_col, containing the central values (typically these are Re estimates)
CI_up, upper limit of the confidence interval
CI_down, the lower limit of the confidence interval