civic_audit: Generate a reproducible audit trail

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

Generate a reproducible audit trail

Description

Produces a structured JSON audit record for any 'civic_model': provenance metadata (data hash, seed, analyst, timestamp), performance metrics, equity summary, and DataCitizen-Pro annotations.

Usage

civic_audit(
  object,
  metrics = NULL,
  fairness = NULL,
  notes = NULL,
  analyst = NULL,
  path = NULL
)

Arguments

object	A 'civic_model'.
metrics	Named numeric vector from [civic_metrics()] (optional).
fairness	A 'civic_fairness' from [civic_fairness()] (optional).
notes	Character analyst notes (optional).
analyst	Character analyst name (optional).
path	File path to write the JSON (optional).

Value

Invisibly, the JSON character string.

Examples

m <- civic_fit(voted ~ age + education, civic_voting)
trail <- civic_audit(m, analyst = "O. O. Awe", notes = "Baseline")
cat(trail)

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