Canonical data, history and validation

In riskweightedassets: Reproducible Risk-Weighted Asset Calculations

knitr::opts_chunk$set(collapse = TRUE, comment = "#>")

The engine consumes 68 named tables. Each historised record distinguishes business validity from the time at which the information became known. The official snapshot is therefore reproducible after subsequent corrections.

library(riskweightedassets)
tables <- generate_synthetic_tables()
head(tables$run_config[c("config_key", "config_value")])
head(tables$exposure_lot[c("record_id", "business_key", "valid_from",
                           "known_from", "is_official")])

Workbook validation returns structured messages rather than prose-only logs.

report <- validate_dataset("path/to/dataset")
issues <- as.data.frame(report)
subset(issues, severity == "ERROR")

Technical validation cannot decide institution-specific scope, permissions or legal interpretation. Those remain governed inputs.

riskweightedassets documentation built on Sept. 24, 2026, 5:12 p.m.