civic_scorecard: Generate a full civic accountability scorecard

In civic.icarm: Interpretable Civic-Accountable and Responsible Machine Learning

Generate a full civic accountability scorecard

Description

Synthesises model provenance, interpretability rating, performance, and equity into a printed civic accountability scorecard, with an optional JSON output. Works for all task types.

Usage

civic_scorecard(
  object,
  test_data,
  outcome,
  protected = NULL,
  positive = NULL,
  analyst = NULL,
  project = "civic.icarm",
  path = NULL
)

Arguments

object	A 'civic_model'.
test_data	Data frame of held-out test data.
outcome	Character. Outcome column name.
protected	Character. Protected attribute column (optional).
positive	Positive class (binary only).
analyst	Character analyst name.
project	Character project name.
path	Optional file path for JSON output.

Value

Invisibly, a named list (the scorecard structure).

Examples

splits <- civic_split(civic_voting, stratify = "voted")
m <- civic_fit(voted ~ age + education + political_interest, splits$train)
civic_scorecard(m, splits$test, outcome = "voted",
                protected = "gender", positive = "yes",
                project = "DataCitizen-Pro")

civic.icarm documentation built on June 18, 2026, 1:06 a.m.