Skip to contents

Changelog

Source: NEWS.md

rTrafa 0.1.0

Initial CRAN release.

Features

  • API connection: trafa_available() performs a lightweight connectivity check, used to guard examples and tests.
  • Product discovery: get_products() lists all statistical products (datasets) available on the Trafa API.
  • Structure inspection: get_measures() and get_dimensions() retrieve the measures (KPIs) and filterable dimensions for a given product, including hierarchy metadata and dimension validation.
  • Data retrieval: get_data() downloads data using pipe-delimited query syntax, with automatic parsing and optional simplification (human-readable _label columns alongside raw codes).
  • Query workflow: prepare_query() validates selections against the structure endpoint before hitting the data endpoint; compose_structure_query() and compose_data_query() expose programmatic access to the raw URL builders.
  • Entity operations: each entity type (product, measure, dimension) supports a consistent family of *_search(), *_describe(), *_extract_ids() / *_extract_names() and *_minimize() helpers for piped exploration.
  • Filter shortcuts: dimension_values() surfaces Trafa’s server-side shortcuts (senaste, forra) alongside regular values, so queries can always reach the latest period without hardcoding years.
  • Data helpers: data_minimize() drops monotonous columns; data_legend() generates a source caption suitable for ggplot2::labs(caption = ...), with lang, omit_varname and omit_desc arguments for fine-grained control over what to show.
  • Persistent caching: trafa_cache_dir() and trafa_clear_cache() manage cached API responses via tools::R_user_dir().
  • HTTP resilience: automatic retry with exponential backoff for transient errors.
  • Offline-safe: all examples and vignettes are guarded by trafa_available() and draw on pre-cached API data stored in R/sysdata.rda, so package builds and tests do not require network access.

Documentation

  • Quick start vignette (a-quickstart-rtrafa): five-step walk-through from product discovery to a plotted time series.
  • Introduction vignette (introduction-to-rtrafa): covers the four-level data model (product → measure → dimension → value), hierarchies, filter shortcuts, dimension validation, prepared queries, and three worked ggplot2 examples.
  • Vignette plots convert the ar (year) column to Date before plotting and use scale_x_date() so axis breaks land on whole years — a pattern the sibling packages rKolada and pixieweb share.
  • README and vignettes cross-link to the sibling packages rKolada (Swedish municipal and regional KPIs) and pixieweb (PX-Web APIs for Nordic statistics agencies).