Package {geoDeltaAudit}

Title: Quantifying Variable Change Induced by Administrative Boundary Transformations
Version: 0.1.0
Description: Tools for auditing how analytic variables change when data are transformed across administrative boundary systems. The package is agnostic to data source, variable type, and administrative geography, and is designed to quantify transformation-induced change without attributing blame to any specific boundary definition or allocation scheme.
License: MIT + file LICENSE
URL: https://github.com/phinnphace/geoDeltaAudit
BugReports: https://github.com/phinnphace/geoDeltaAudit/issues
Depends: R (≥ 4.1.0)
Imports: dplyr, janitor, rlang, stringr, tibble,
Suggests: knitr, readr, rmarkdown, testthat (≥ 3.0.0)
VignetteBuilder: knitr
Encoding: UTF-8
RoxygenNote: 7.3.3
NeedsCompilation: no
Packaged: 2026-05-09 15:11:16 UTC; phinnmarkson
Author: Phinn Markson [aut, cre]
Maintainer: Phinn Markson <markson.2@osu.edu>
Repository: CRAN
Date/Publication: 2026-05-13 19:10:07 UTC

Audit a sequence of geographic transformations

Description

testthat::skip("Integration test: run manually (slow / uses real data).") Computes delta_x(VAR) between a baseline and a transformed result while returning diagnostics.

Usage

audit_transform(
  df,
  geo_col,
  var_col,
  steps,
  baseline_filter = NULL,
  target_id = NULL
)

Arguments

df

Input data frame.

geo_col

Column containing geography IDs.

var_col

Column containing the variable of interest.

steps

A list of step functions created by step_* helpers.

baseline_filter

Optional function(df) -> filtered df defining baseline membership.

target_id

Optional target ID to extract after final step (e.g., "27053").

Value

An object of class audit_result.

Normalize messy geography headers to standard names

Description

Normalize messy geography headers to standard names

Usage

clean_geo_headers(df, map, keep)

Arguments

df

A data frame with geography columns.

map

Named character vector: names are standardized outputs, values are regex patterns of accepted input names.

keep

Character vector of standardized columns to keep.

Value

A tibble with standardized names.

Prepare HUD ZIP-to-County crosswalk

Description

Standardizes HUD crosswalk fields and enforces string IDs.

Usage

prep_hud_crosswalk(data, ratio_col = "TOT_RATIO")

Arguments

data

Raw HUD crosswalk data frame.

ratio_col

Which HUD ratio to use (default: "TOT_RATIO").

Value

Tibble with columns: zip, county, tot_ratio.

Step: ZCTA -> ZIP using equal-share allocation

Description

Given an association table mapping ZCTAs to ZIPs, allocate each ZCTA's values equally across its associated ZIPs.

Usage

step_zcta_to_zip_equal(assoc, zcta_col = "zcta", zip_col = "zip")

Arguments

assoc

A data frame containing ZCTA-ZIP associations.

zcta_col

Column name in assoc containing ZCTA IDs (ignored if clean_geo_headers matches).

zip_col

Column name in assoc containing ZIP IDs (ignored if clean_geo_headers matches).

Value

A step function suitable for audit_transform().

Step: ZIP -> COUNTY using HUD TOT_RATIO

Description

Allocate ZIP-level values to counties using HUD's TOT_RATIO weights.

Usage

step_zip_to_county_totratio(
  hud,
  zip_col = "zip",
  county_col = "county",
  weight_col = "tot_ratio"
)

Arguments

hud

A data frame containing ZIP-to-county weights.

zip_col

Column name for ZIP (kept for API symmetry; cleaning is robust).

county_col

Column name for county (FIPS) (kept for API symmetry).

weight_col

Column name for the weight (default "tot_ratio") (kept for API symmetry).

Value

A step function suitable for audit_transform().