Skip to contents

Changelog

Source: NEWS.md

nycOpenData 0.2.2

Major improvements

This release continues the transition to a catalog-driven workflow for accessing NYC Open Data datasets in R, emphasizing transparency, reproducibility, and streamlined API access.

  • nyc_list_datasets() now serves as the primary entry point for discovering datasets available through the live NYC Open Data catalog.
  • nyc_pull_dataset() provides a unified interface for downloading datasets using either human-readable catalog keys or Socrata dataset UIDs.
  • Dataset keys are generated dynamically using janitor::make_clean_names() based on live catalog metadata.
  • Expanded support for flexible dataset discovery and retrieval without requiring dedicated dataset wrapper functions.

New features

  • Added support for flexible dataset selection using either:
    • human-readable catalog keys (e.g., "motor_vehicle_collisions_crashes")
    • or stable Socrata dataset UIDs (e.g., "h9gi-nx95")
  • Added support for date filtering using date, from, and to arguments with user-specified date_field.
  • Improved filtering support with vectorized filters translated into SQL-style IN clauses.
  • Improved support for direct Socrata endpoint access through nyc_any_dataset().

Documentation

  • Substantially expanded README documentation to clarify:
    • package architecture
    • Socrata API workflows
    • SoQL usage
    • dataset keys versus UIDs
    • catalog-driven dataset retrieval
    • type coercion behavior
  • Improved function documentation for:
  • Updated examples to demonstrate:
    • realistic catalog search workflows
    • filtering
    • date-based queries
    • direct endpoint access
  • Revised vignettes to align with the new catalog-driven architecture and remove outdated wrapper-function references.
  • Added additional development and testing documentation for vcr and webmockr.

Internal improvements

  • Refactored internal request handling to support dynamic catalog lookup.
  • Improved validation for dataset identifiers, filters, and date-filtering parameters.
  • Improved handling of conservative heuristic-based type coercion.
  • Reduced reliance on dedicated dataset wrapper functions in favor of a unified interface.

Maintenance

  • Updated package citation information and version consistency across package documentation.
  • Continued improvements to API reliability, documentation clarity, and internal utilities.