Report target-population projection evidence

`margot_target_projection_report()` compares the realised source sample with the registered target population before and after projection weighting. It reports aggregate balance, projection-weight concentration, and source representation across registered strata. The function returns evidence and provenance without an accept-or-reject classification.

Usage

margot_target_projection_report(
  source,
  target,
  variables,
  projection_weights = NULL,
  target_weights = NULL,
  strata = NULL,
  harmonisation = NULL,
  projection_model = NULL,
  uncertainty = NULL,
  top_shares = c(0.01, 0.05, 0.1)
)

Arguments

source

A source-sample data frame.

target

A target-population data frame or aggregate target microdata.

variables

Character vector naming the registered projection variables present in both data frames.

projection_weights

Optional non-negative source-sample projection weights. Equal weights represent an unweighted projection.

target_weights

Optional non-negative target-population weights. Equal weights are used when this argument is `NULL`.

strata

Optional subset of `variables` defining the registered overlap strata. The report describes target strata that lack source representation.

harmonisation

Optional aggregate record of the source-to-target harmonisation rules.

projection_model

Optional aggregate description of the projection model or supplied-weight construction.

uncertainty

Optional aggregate description of projection-weight or target-margin uncertainty.

top_shares

Fractions used for projection-weight concentration summaries.

Value

An object of class `margot_target_projection_report` containing a variable-level balance table, stratum representation table, projection-weight summary, concentration table, and provenance. The return value contains no participant-level rows.