Changelog

Source: NEWS.md

emburden 0.6.0

New Features

• Housing Characteristics Analysis: Preserve granular housing dimension columns (TEN, TEN-YBL6, TEN-BLD, TEN-HFL) through cohort data aggregation, enabling detailed analysis of energy burden by tenure, building age, building type, and heating fuel
• Implement comprehensive dev/staging/main branching strategy

Bug Fixes

• Exclude backtick-wrapped strings from BREAKING CHANGE detection
• Move LaTeX package installation after R dependencies setup
• Add LaTeX package installation for vignette builds
• Exclude deprecated workflows from actionlint validation
• Exclude heredoc-containing workflows from actionlint
• Resolve YAML syntax errors in workflow heredocs
• Replace automated CRAN submission with manual instructions
• Use rhub::submit_cran for automated CRAN submission
• Pass package directory to submit_cran, not tarball path
• Remove email parameter from devtools::submit_cran()
• Use r-lib/actions to install devtools for CRAN submission
• Use build_args in rcmdcheck for proper vignette compaction
• Add –compact-vignettes=both to resolve PDF size warning

Enhancements

• Documentation: Enhanced load_cohort_data() documentation with detailed descriptions of housing dimension columns and examples showing housing characteristic analysis
• Vignettes: Added comprehensive housing characteristics analysis section to getting-started vignette with examples of tenure, building age, building type, and heating fuel analysis
• Testing: Added 20+ comprehensive tests for housing dimension preservation through data aggregation pipeline
• Citations: Updated inst/CITATION to reflect published Nature Communications paper (Scheier & Kittner, 2022) with DOI
• Add comprehensive branching strategy documentation
• Add status badges to README
• Update CONTRIBUTING.md for new branching workflow
• Add CRAN-SUBMISSION to .Rbuildignore

Other Changes

• Release: CRAN workflow improvements and semantic versioning
• config: Limit GitHub Actions to only main branch

---

emburden 0.5.19

New Features

• Add tail -1 safety check for tarball extraction
• Add comprehensive build diagnostics to publish workflow

Bug Fixes

• Use working grep pattern for tarball extraction
• Remove fallback tarball extraction, fail explicitly instead

---

emburden 0.5.18

New Features

• Add full validation pipeline to public repo before CRAN submission
• Implement multi-stage CRAN submission workflow

Bug Fixes

• Add validate-cran to dry-run-prod needs for output access
• Ensure numeric values in auto-merge check loop

Enhancements

• Move CRAN validation pipeline to private repo
• Convert to sequential CRAN submission pipeline

---

emburden 0.5.17

Bug Fixes

• Prevent duplicate tag creation in release workflow
• Skip CRAN workflow gracefully on private repo
• Improve CRAN workflow job dependencies and output visibility

Other Changes

• hotfix: Skip vignette building in publish-to-public CRAN validation
• Fix CRAN R dependency and auto-release workflow
• Fix public repo releases and add auto-increment version feature
• Update auto-release.yml
• Update deployment workflow

---

emburden 0.5.16

New Features

• Trigger auto-merge via tag push from release script
• Add auto-merge for version bumps and improve CRAN workflows

Bug Fixes

• Add wait loop for CI checks in auto-merge workflow
• Update R dependency to 4.1.0 for native pipe support
• Prevent duplicate tag creation in release workflow
• Skip CRAN workflow gracefully on private repo
• Improve CRAN workflow job dependencies and output visibility

Other Changes

• Fix public repo releases and add auto-increment version feature
• Update auto-release.yml
• Update deployment workflow

---

emburden 0.5.15

New Features

• Add auto-increment version feature to release script

Bug Fixes

• Enable workflow_dispatch trigger to create releases on public repo
• Prevent duplicate tag creation in release workflow
• Skip CRAN workflow gracefully on private repo
• Improve CRAN workflow job dependencies and output visibility

Other Changes

• Update auto-release.yml
• Update deployment workflow

---

emburden 0.5.99

New Features

• Complete workflow_call implementation for all triggers

Bug Fixes

• Use current branch ref for publish-to-public trigger
• Add workflow_dispatch inputs to publish-to-public
• Prevent duplicate tag creation in release workflow
• Skip CRAN workflow gracefully on private repo
• Improve CRAN workflow job dependencies and output visibility

Other Changes

• Update auto-release.yml
• Update deployment workflow

---

emburden 0.5.14

Changes

New Features

• (Add new features here)

Bug Fixes

• (Add bug fixes here)

Enhancements

• (Add enhancements here)

---

emburden 0.5.13

Changes

1. Removed redundant tag creation - Release script no longer creates tags; workflow handles it after merge
2. Remote tag checking - Workflow now checks both local and remote tags to prevent conflicts
3. Automatic PR creation - Bonus feature that extracts NEWS.md content for PR bodies

---

# emburden 0.5.12

## Changes

### New Features

• Release Automation: Added automatic PR creation to release-version.sh script
  • Automatically creates or finds existing PR after pushing version tag
  • Extracts and includes NEWS.md content for current version in PR body
  • Displays PR URL in final summary output
  • Includes comprehensive release automation guide (.dev/RELEASE-AUTOMATION-GUIDE.md)

### Bug Fixes

• CRAN Workflow: Fixed workflow to skip gracefully on private repository instead of failing
  • Changed from step-level exit 1 to job-level if: github.repository == 'ericscheier/emburden' condition
  • Ensures clean skip (not failure) when running on ScheierVentures/emburden
  • Only executes on public repository for CRAN submissions

### Enhancements

• Developer Experience: Release script now handles full workflow from version bump to PR creation in 9 automated steps

# emburden 0.5.11

## Changes

### Bug Fixes

• CRAN Workflow: Improved job dependencies and output visibility
  • Fixed submit-to-cran job to only run when validate-cran succeeds (added if: success() condition)
  • Changed R CMD build to use tee to show all output while capturing tarball filename
  • Added error handling for tarball name extraction failures
  • Prevents “out of order” issues where jobs try to download artifacts that were never uploaded
• Git LFS: Completely removed Git LFS to resolve budget exceeded errors
  • Deleted .gitattributes (LFS configuration)
  • Removed LFS config from auto-tag workflow
  • Updated .gitignore to exclude R package build artifacts (Meta/, doc/)
• CI/CD: Added repository check to CRAN Release workflow
  • Only runs on public repository (ericscheier/emburden)
  • Prevents wasted CI minutes on private repository
  • CRAN submissions now only come from public repo

---

emburden 0.5.10

Workflow Organization

This release reorganizes the CRAN release workflows between private and public repositories.

Changes

• Workflow documentation:
  • Added comprehensive deployment guide for public repository CRAN workflow
  • Updated workflow README to clarify automatic triggering and approval gates
  • Prepared controlled-release workflow for public repository deployment
• Repository architecture:
  • Private repo focuses on fast GitHub releases via auto-release
  • Public repo handles CRAN validation with automatic triggering and dual approval gates
  • Eliminates workflow conflicts by sequential execution

---

emburden 0.5.9

CI/CD Improvements

This release focuses on improving the robustness and reliability of the CI/CD pipeline.

New Features

• Workflow validation system:
  • Added .github/workflows/validate-workflows.yml for PR validation
  • Pre-tag validation gate in auto-tag workflow prevents creating tags when workflow files have syntax errors
  • Manual validation script .dev/validate-workflows.sh for local checks
  • Prevents misordering of workflow fixes and version bumps

Bug Fixes

• Fixed Windows CI TinyTeX failures:
  • Added tlmgr_update() before installing LaTeX packages
  • Resolves “outdated CTAN mirror” errors on Windows runners
  • Ensures consistent vignette building across all platforms (Windows, macOS, Linux)

---

emburden 0.5.8

CRAN Automation and Submission

This release introduces comprehensive automation for CRAN submissions.

New Features

• Automated CRAN submission pipeline:
  • Multi-layer validation (local → GitHub Actions → Win-builder → manual approval → auto-submit)
  • Win-builder integration for Windows testing
  • Manual approval gate via GitHub environment
  • Automatic submission using devtools::submit_cran()
• Pre-tag validation script (.dev/pre-tag-cran-check.R):
  • Local validation before creating version tags
  • Comprehensive CRAN checks with --as-cran flag
  • Optional Win-builder submission
  • Version consistency validation
• Complete workflow documentation (.dev/CRAN-SUBMISSION-GUIDE.md):
  • Full CRAN submission process guide
  • Multi-repository setup explanation
  • Troubleshooting tips and best practices

Bug Fixes

• Fixed R-CMD-check badge URL in README (.yaml → .yml)

---

emburden 0.5.7

CRAN Readiness - Final Fixes

This patch release completes CRAN readiness with final compliance fixes.

Bug Fixes

• Package build exclusions: Excluded data/zenodo-upload-nationwide/ directory from package tarball (fixes CRAN data directory WARNING)
• Spelling whitelist: Added inst/WORDLIST with 85 technical terms and acronyms to prevent false-positive spelling errors
• Public repository sync: Fixed publish-to-public workflow to properly remove private-only workflow files before syncing to public repository

---

emburden 0.5.6

CRAN Quality-of-Life Improvements

This release focuses on CRAN compliance and automation improvements.

Enhancements

• Auto-release workflow: Automated GitHub release creation on version tags
• CRAN compliance improvements:
  • Added Language: en-US field to DESCRIPTION
  • Added jsonlite to Suggests (used in tests)
  • Replaced all non-ASCII Unicode characters with escape sequences
  • Added missing global variable bindings (AMI150, AMI68)

Bug Fixes

• Fixed auto-release workflow heredoc syntax issue
• All Unicode characters now use \uxxxx escape format for portability

---

emburden 0.5.5

Data Integrity Fix - New Zenodo Record

This patch release deploys corrected datasets to a new Zenodo record to ensure data integrity.

Bug Fixes

• New Zenodo record with verified correct data
  • Deployed new Zenodo record 10.5281/zenodo.17656637
  • Updated MD5 checksums to match re-uploaded files with verified correct data
  • Verified FPL 2022 checksum: 767f2ff27193116f61e893999eb8bcf1
  • Impact: Ensures users download validated, correct data for all 4 datasets

---

emburden 0.5.4

Critical Bugfix - Zenodo MD5 Checksums

This patch release fixes incorrect MD5 checksums that caused data loading failures.

Bug Fixes

• Fixed MD5 checksums for Zenodo downloads
  • Corrected AMI 2022 checksum: cc847d89119a374bede6ee7f41060506
  • Corrected AMI 2018 checksum: 4941e3566daec1badc53eb44f34d95a8
  • Corrected FPL 2018 checksum: 85ef6b7b4de244e80ff700f3d5becf3a
  • Updated file sizes to match actual generated files
  • Impact: Previously, 3 out of 4 datasets failed checksum verification and fell back to cached/OpenEI data, causing incorrect data comparisons (e.g., 2018 and 2022 appearing identical)

---

emburden 0.5.3

Zenodo Integration - US Nationwide Datasets

This patch release enables Zenodo downloads for US nationwide datasets with improved reliability and performance.

Data Infrastructure

• Enabled Zenodo downloads for US nationwide datasets (PR #35)
  • Deployed Zenodo record 10.5281/zenodo.17653871 with 4 datasets
  • AMI cohorts 2022 (499,234 records, 51 states)
  • FPL cohorts 2022 (416,054 records, 51 states)
  • AMI cohorts 2018 (361,095 records, 51 states)
  • FPL cohorts 2018 (361,085 records, 51 states)
  • Updated MD5 checksums for all datasets
  • Removed temporary Zenodo bypass code

Bug Fixes

• Fixed test mocking for database fallback (PR #35)
  • Database fallback test now properly mocks all download sources
  • Added mock for download_lead_data() to prevent OpenEI fallback
  • Added mock for detect_database_corruption() to allow test data

Testing

• All 614 tests passing across 7 platforms
• Clean R CMD check: 0 ERRORS, 0 FAILURES

---

emburden 0.5.2

CRAN Submission Fix - LaTeX Compatibility

This patch release fixes a LaTeX compatibility issue blocking CRAN submission.

Bug Fixes

• Fixed LaTeX Unicode error in documentation (PR #32)
  • Replaced Unicode ≥ character (U+2265) with LaTeX-compatible \eqn{\ge} macro
  • Fixed in R/energy_ratios.R documentation for ner_func() function
  • All R CMD check tests passing with 0 ERRORS

CRAN Readiness

• Clean R CMD check results: 0 ERRORS, 1 WARNING (qpdf - non-critical), 3 NOTEs (all acceptable)
• All 614 tests passing across 7 platforms (ubuntu, windows, macos, multiple R versions)
• Package ready for CRAN submission

---

emburden 0.5.1

Critical Data Fix - Corrected Zenodo Repository

This patch release fixes critical data corruption in the v0.5.0 Zenodo repository.

Bug Fixes

• Fixed corrupted Zenodo data (PR #28)
  • v0.5.0 Zenodo record (17605603) contained incorrect FPL data files
  • FPL files only included NC state data (52MB) instead of full nationwide data (306MB)
  • AMI files were correct (nationwide data, 148MB)
  • New Zenodo record (17613104) uploaded with all 4 corrected datasets
  • All datasets now contain complete US nationwide data (51 states, ~73K census tracts)
• Updated Zenodo configuration
  • New concept DOI: 10.5281/zenodo.17613103
  • New version DOI: 10.5281/zenodo.17613104
  • Updated all file URLs and MD5 checksums in R/zenodo.R
  • Updated test patterns to accept new Zenodo API endpoint format

Verified Data Integrity

All 4 nationwide datasets verified and working correctly: - lead_ami_cohorts_2022_us.csv.gz - 148 MB ✓ - lead_fpl_cohorts_2022_us.csv.gz - 305 MB ✓ - lead_ami_cohorts_2018_us.csv.gz - 148 MB ✓ - lead_fpl_cohorts_2018_us.csv.gz - 305 MB ✓

All tests passing (614 tests, 0 failures).

---

emburden 0.5.0

CRAN Submission Ready - Nationwide Energy Burden Analysis

This major release marks the completion of the nationwide expansion and prepares the package for CRAN submission. The package now comprehensively showcases nationwide US capability across all documentation, with 648 tests passing and clean R CMD check results.

Nationwide Expansion Complete

• Full nationwide focus achieved across all documentation
  • README features nationwide data from introduction: “All 51 US states…2.3+ million records”
  • All function examples demonstrate single-state → multi-state → nationwide progression
  • Both vignettes showcase nationwide capability alongside learning examples
  • Dual focus strategy: NC examples for learning (small, fast), nationwide for production use
• Comprehensive test coverage validates nationwide functionality
  • 648 tests passing (0 failures)
  • Multi-state regional filtering (Southeast, top 10 states, cross-regional)
  • Data integrity validation across all 51 states
  • All major US regions tested (Northeast, Southeast, Midwest, Southwest, West)
• CRAN readiness verified
  • R CMD check: 0 errors, 1 acceptable warning (qpdf), 1 acceptable note (httptest2)
  • Package size: Under 5MB CRAN limit (~1.9MB)
  • Multi-platform CI validation (macOS, Windows, Ubuntu × 5 R versions)
  • External data hosting on Zenodo (DOI: 10.5281/zenodo.17605603)

Documentation Enhancements

• Nationwide vignette content
  • vignettes/getting-started.Rmd: Comprehensive nationwide examples (v0.4.10)
  • vignettes/jss-emburden.Rmd: Nationwide data availability note added
  • Performance guidance for large dataset queries (30-120 seconds, ~500MB RAM)
  • Metadata discovery functions showcased (list_states(), list_income_brackets(), etc.)
• Language cleanup
  • Removed “proof of concept” references from documentation
  • Professional, production-ready messaging throughout
  • Clear data coverage statements: 2.3M+ household records, ~73k census tracts, all 51 states

Data Infrastructure

• Zenodo data hosting (established in v0.4.7-0.4.8)
  • 4 nationwide datasets published (AMI/FPL 2018/2022, 307 MB compressed)
  • MD5 checksum verification for data integrity
  • Automatic download cascade: Database → CSV → Zenodo → OpenEI fallback
  • Package stays under CRAN 5MB limit

Package Quality Metrics

• Test coverage: 648 comprehensive tests
  • 99 multi-state and nationwide tests
  • 48 metadata discovery tests
  • 62 Zenodo integration tests
  • Complete data loader and comparison function coverage
• CI/CD infrastructure
  • Multi-platform R CMD check (5 environments)
  • Test coverage reporting
  • Automated release workflow on version bumps
  • Pre-commit and pre-push hooks for local validation

Breaking changes: None. All existing NC-focused code continues to work. Nationwide capability is additive.

Next milestone: CRAN submission! 🚀

emburden 0.4.9

Documentation Transition & Infrastructure

• NC→Nationwide transition (Phase 1): Package documentation now showcases nationwide US capability
  • Updated README.md with multi-state and nationwide examples alongside NC examples
  • Updated all function examples (compare_energy_burden(), load_cohort_data(), load_census_tract_data())
  • Added test validating all 51 US states are supported (614 tests passing)
  • Data coverage: 2.3M household cohort records, ~73k census tracts, all 51 states
  • Follows “dual focus” strategy: NC examples for learning, nationwide examples for production use
  • See .dev/NC-TO-NATIONWIDE-TRANSITION.md for comprehensive transition plan
• pkgdown build fix: Resolved recurring CI failure
  • Changed backup_db() and clear_test_environment() from @export to @keywords internal
  • Added pkgdown reference index check to pre-commit hook to prevent recurrence
  • Hook provides helpful hints about @export vs @keywords internal

No breaking changes: All NC-focused examples continue to work. Nationwide data access is additive.

emburden 0.4.8

Database Protection & Testing Infrastructure

• Production database protection to prevent accidental data loss:
  • New R/database-helpers.R module with safe database operations
  • delete_db() requires explicit confirm = TRUE for production database
  • backup_db() creates timestamped backups before risky operations
  • clear_test_environment() safely clears only test data
  • Separate test (emburden_test_db.sqlite) and production (emburden_db.sqlite) databases
  • All database helpers fully documented with roxygen2
• Zenodo integration completed with NATIONWIDE data publication:
  • Updated R/zenodo.R with published Zenodo record (DOI: 10.5281/zenodo.17605603)
  • 4 NATIONWIDE datasets uploaded (AMI/FPL 2018/2022, 307 MB compressed, all 51 US states)
  • 2.3+ million cohort records covering ~73,000 census tracts
  • All download functions now use real Zenodo URLs
  • MD5 checksum verification for all downloads
  • Automated Zenodo upload and R code update scripts
  • Comprehensive test suite (48 new metadata tests + 62 zenodo tests = 604 total tests)
• Comprehensive test coverage for Zenodo infrastructure:
  • tests/testthat/test-zenodo-integration.R: Configuration and database protection tests
  • tests/testthat/test-zenodo-download.R: Download functionality tests
  • Fixed tests/testthat/test-data-loaders.R for Zenodo download cascade
  • All 556 tests passing (0 failures, 3 expected offline skips)
• Development tools for data management:
  • .dev/upload-to-zenodo-nationwide.sh: Automated nationwide Zenodo upload via REST API
  • .dev/update-zenodo-config.R: Auto-update R/zenodo.R from upload output
  • .dev/prepare-zenodo-data-nationwide.R: Script for preparing all 51 states
  • .dev/NC-TO-NATIONWIDE-TRANSITION.md: Comprehensive transition plan
  • .dev/TEST_ZENODO_DOWNLOAD.md: Complete testing guide
  • Updated .gitignore for build artifacts
• Metadata discovery functions with comprehensive tests:
  • list_states(): Returns all 51 US state abbreviations
  • list_income_brackets(): Income brackets by dataset/vintage
  • list_cohort_columns(): Column names and descriptions
  • get_dataset_info(): Complete dataset metadata
  • 48 new tests in tests/testthat/test-metadata.R

Testing workflow: Safe TDD workflow established with test database isolation

Next steps: Transition documentation from NC-focused to nationwide (see .dev/NC-TO-NATIONWIDE-TRANSITION.md), ready for CRAN submission

emburden 0.4.7

Data Hosting Infrastructure

• Implemented Zenodo data hosting with OpenEI fallback:
  • New R/zenodo.R module for downloading from Zenodo repository
  • Faster downloads via Zenodo CDN vs OpenEI
  • MD5 checksum verification for data integrity
  • Gzip decompression support for smaller downloads
  • Automatic fallback to OpenEI if Zenodo unavailable
• Updated download cascade in load_cohort_data() and load_census_tract_data():
  1. Database (SQLite) - fastest, local
  2. CSV (cached files) - fast, local
  3. Zenodo (NEW!) - faster, more reliable
  4. OpenEI (fallback) - original source
• Added maintainer documentation: .dev/ZENODO_UPLOAD_GUIDE.md
  • Complete workflow for preparing and uploading datasets
  • Compression and checksum procedures
  • Testing and versioning guidelines

Benefits: Nationwide data testing ready, package stays under CRAN 5MB limit (currently 1.9MB), improved download reliability

Next steps: Upload processed datasets to Zenodo, update DOI configuration, ready for CRAN submission

emburden 0.4.6

CRAN Preparation Fixes

• Fixed R CMD check WARNINGs and NOTEs:
  • Excluded JSON data files (579MB) from package build via .Rbuildignore
  • Excluded top-level presentation/poster files (25+ non-standard files)
  • Added vignette metadata to jss-emburden.Rmd (VignetteEngine, VignetteIndexEntry)
  • Package now builds cleanly under 5MB for CRAN submission
• Next steps for CRAN: Implement Zenodo data hosting to fully separate data from methods package

emburden 0.4.5

New Features

• Metadata discovery functions for easier data exploration:
  • list_income_brackets(dataset, vintage): Show available income brackets
  • list_states(): Show all 51 available state abbreviations
  • list_cohort_columns(dataset, vintage): Show column names, descriptions, and data types
  • get_dataset_info(): Show metadata about all available datasets
  • Enables programmatic discovery of data structure

emburden 0.4.4

Breaking Changes

• Parameter reordering in compare_energy_burden(): group_by now comes before counties
  • New order: compare_energy_burden(dataset, states, group_by, counties, ...)
  • Makes intuitive syntax work: compare_energy_burden('fpl', 'NC', 'income_bracket')
  • Named parameters unaffected: compare_energy_burden(dataset='fpl', counties=c('Orange'))

New Features

• Dynamic grouping in compare_energy_burden(): group_by now accepts custom column names
  • Use keywords: “income_bracket”, “state”, “none” (as before)
  • OR custom columns: group_by = "geoid" for tract-level comparison
  • OR multiple columns: group_by = c("state_abbr", "income_bracket")
  • Enables flexible analysis patterns for full USA data

emburden 0.4.3

New Features

• Dynamic filtering in load_cohort_data(): Now accepts ... parameter for flexible filtering
  • Filter by any column using tidyverse syntax
  • Example: load_cohort_data("ami", states = "NC", households > 100, total_income > 50000)
  • Complements existing states, counties, income_brackets parameters
  • First step toward full USA data package architecture

emburden 0.4.2

Bug Fixes

• Fixed confusing warnings when using compare_energy_burden('fpl', 'NC', 'income_bracket')
• Function now silently handles common mistake of passing ‘income_bracket’, ‘state’, or ‘none’ as counties argument
• Eliminates “County name ‘income_bracket’ not found” warnings while maintaining correct behavior

Improvements

• Improved documentation with clearer examples distinguishing between group_by and counties parameters

emburden 0.4.1

Improvements

• Updated contact email from eric.scheier@gmail.com to eric@scheier.org across all documentation

emburden 0.4.0

New Features

Fully Automated Release Workflow

• Zero-touch releases: GitHub releases now created automatically when version bumps are merged
  • Detects DESCRIPTION version changes automatically
  • Runs all quality checks (R CMD check, tests, coverage)
  • Generates release notes from NEWS.md
  • Creates git tags and GitHub releases with package tarball
  • No manual intervention required!
• Workflow: Simply bump version in DESCRIPTION, update NEWS.md, merge PR → release happens automatically

emburden 0.3.0

Major Improvements

OpenEI Data Pipeline Fix (Critical)

• Fixed critical bug where MVP demo compare_energy_burden('fpl', 'NC', 'income_bracket') failed on fresh installs
  • Root cause: Raw OpenEI 2022 FPL data wasn’t being processed correctly
  • OpenEI data uses period-based columns (HINCP.UNITS) not asterisk-based (HINCP*UNITS)
  • Raw data has ~588k rows (one per housing characteristic combination) requiring aggregation
• New data processing pipeline:
  • Added aggregate_cohort_data() function to aggregate raw data by census tract × income bracket
  • Updated detection logic to recognize both .UNITS and *UNITS column formats
  • Enhanced standardize_cohort_columns() to handle both FPL150 (2022) and FPL15 (2018)
  • Reduces 588k rows → ~3.6k cohort records for NC
• Result: Fresh installations now work perfectly - download from OpenEI → aggregate → standardize → ready!

Orange County Sample Data

• NEW: Bundled sample data for instant demos and testing (94 KB)
  • data(orange_county_sample) - No download required!
  • Includes 4 datasets: fpl_2018, fpl_2022, ami_2018, ami_2022
  • 749 records across 42 census tracts (Orange County, NC)
  • Perfect for vignettes, examples, and quick analysis
  • Shows real data: 16.3% energy burden for lowest income vs 1.0% for highest

Package Infrastructure

• Renamed all internal references: emrgi → emburden for consistency
  • find_emrgi_db() → find_emburden_db()
  • Database filename: emrgi_db.sqlite → emburden_db.sqlite
• Release automation:
  • Added .dev/RELEASE-PROCESS.md - Comprehensive release workflow guide
  • Added .dev/create-release-tag.R - Automated release tagging script

Documentation

• Updated README with Orange County sample data section
• Added comprehensive documentation for orange_county_sample
• All examples now work out of the box with bundled sample data

Testing

• All 494 tests pass
• Verified OpenEI download and processing pipeline with real data
• Tested sample data access and analysis

emburden 0.2.0

New Features

JSS Manuscript Vignette

• Added Journal of Statistical Software (JSS) manuscript as package vignette
  • vignettes/jss-emburden.Rmd - Complete JSS article format
  • Demonstrates package usage with reproducible examples
  • Includes bibliography and proper JSS formatting
  • Test suite ensures vignette builds correctly in CI
• Created manuscript development infrastructure
  • research/manuscripts/jss-draft/ - LaTeX build output
  • research/manuscripts/build-jss.R - Build script for PDF generation
  • Separate from vignettes for flexible editing workflow

Enhanced Temporal Comparison

• Prominently featured compare_energy_burden() function across all documentation
  • README now includes temporal comparison section (Example 5)
  • Getting Started vignette has comprehensive temporal comparison section
  • JSS vignette demonstrates function instead of manual calculations
  • Replaces 37-line manual comparison with elegant 12-line function call

Bug Fixes

• Fixed FPL (Federal Poverty Line) data loading (#15)
  • Added validation to skip files with missing or all-NA income_bracket columns
  • Loader now properly falls through to raw OpenEI files with complete data
  • Prevents “Element income_bracket doesn’t exist” errors

Documentation Improvements

• Emphasized compare_energy_burden() usage across 7 files
  • README.md - Added temporal comparison section
  • vignettes/jss-emburden.Rmd - Replaced manual code with function call
  • vignettes/getting-started.Rmd - Added comprehensive section
  • analysis/scripts/nc_comparison_for_email.R - Complete rewrite (179→144 lines)
  • data-raw/README.md - Fixed function references
  • research/manuscripts/jss-draft/jss-emburden.Rmd - Updated examples
• Added pkgdown configuration for JSS vignette
  • Vignette appears in website navigation
  • Organized under “Package Documentation” section

Infrastructure

• Added pre-commit hook for running package tests
  • .git/hooks/pre-commit - Runs all 238 tests before each commit
  • Prevents committing broken code
  • Can be bypassed with --no-verify if needed

Internal Changes

• Improved data validation in load_cohort_data()
  • Better handling of incomplete processed CSV files
  • More informative verbose messaging

emburden 0.1.1

Documentation and Infrastructure Improvements

This patch release improves documentation accessibility and workflow infrastructure, with no code changes.

Documentation

• Improved README accessibility and tone
  • Simplified technical language with plain-language explanations
  • Replaced prescriptive language (“WRONG”, “NEVER”) with educational tone (“Recommended”, “Note”)
  • Added concrete examples explaining why simple averaging of ratios fails
• Added complete Nature Communications citation
  • Scheier, E., & Kittner, N. (2022). A measurement strategy to address disparities across household energy burdens
  • Includes BibTeX format for easy reference

Infrastructure

• Changed git author in publish-to-public workflow from “GitHub Actions Bot” to “Eric Scheier”
  • Automated commits now appear as maintainer commits

emburden 0.1.0

Package Release

Initial formal release with package renamed from netenergyequity to emburden for clarity and CRAN compatibility.

This is the first release of the netenergyequity package, providing tools for household energy burden analysis using Net Energy Return methodology.

Core Functionality

• Energy metric calculations
  • energy_burden_func() - Calculate energy burden (S/G)
  • ner_func() - Calculate Net Energy Return (Nh)
  • eroi_func() - Calculate Energy Return on Investment
  • dear_func() - Calculate Disposable Energy-Adjusted Resources
• Statistical analysis
  • calculate_weighted_metrics() - Weighted aggregation with proper Nh methodology
  • Automatic poverty rate calculations below specified thresholds
  • Support for grouped analysis by geographic/demographic categories
• Formatting utilities
  • to_percent(), to_dollar(), to_big() - Publication-ready formatting
  • to_million(), to_billion_dollar() - Compact number formats
  • colorize() - Output-aware color formatting for R Markdown

Package Structure

• Separated package code (R/) from analysis scripts (analysis/)
• Comprehensive documentation with roxygen2
• Test suite with testthat
• Example analysis scripts for NC electric utilities

Known Issues

• roxygen2 documentation generation requires manual NAMESPACE for now
• Large data files (1.1GB+) not included in package distribution
• Some analysis scripts need updating to use package functions

Future Plans

• Add vignettes demonstrating methodology
• Setup pkgdown website
• Configure GitHub Actions for CI/CD
• Consider CRAN submission
• Create companion data package or external data hosting solution