Skip to content

2.0.0 CRAN release

Latest

Choose a tag to compare

@walkerke walkerke released this 20 Oct 16:09

In version 2.0.0:

  • get_estimates() now defaults to Vintage 2025 and supports Puerto Rico: municipio population and age / sex characteristics are available with geography = "county", state = "PR", and Puerto Rico metro areas are included in CBSA and CSA population estimates (#581).
  • get_estimates() now supports the Census Bureau's intercensal estimates with product = "intercensal": use vintage = 2020 for 2010-2020 and vintage = 2010 for 2000-2010. Population totals are available for states, counties, and places, and characteristics by age, sex, race, and Hispanic origin for counties (#629). Population characteristics requests also now support geometry = TRUE and multiple states for counties.
  • New drop_empty argument in get_acs(), get_decennial(), and get_estimates() drops rows with empty geometries (typically water-only areas not in the cartographic boundary files) when geometry = TRUE (#650). These functions also now have an explicit cb argument, which defaults to TRUE (cartographic boundary files); use cb = FALSE for TIGER/Line files (#604).
  • get_estimates() now supports housing unit estimates for 2020 and later with product = "housing" (Vintage 2021 and later) for the US, regions, states, and counties. The Census Bureau publishes these only as Excel tables, so the readxl package is required.
  • get_flows() now defaults to year = 2020, the most recent ACS Migration Flows release with county-to-county flows, and supports the 2017-2021 and 2018-2022 releases, which provide flows between counties and states or world regions (inflows only, with no county subdivision, MSA, or geometry support).
  • get_acs() and get_decennial() now support metropolitan divisions (geography = "metropolitan division") and tribal census tracts (geography = "tribal census tract"), returned for the entire US (#519, #621).
  • New suffix argument in get_acs() sets the estimate and margin of error suffixes for output = "wide", e.g. suffix = c("_est", "_moe"); the default remains c("E", "M") (#600).
  • Cleaned variable labels: get_acs() and get_decennial() have a new labels argument. labels = TRUE names variables from their labels instead of their IDs (e.g. total_male_under_5_years), in tidy or wide output; labels = "concept" adds the concept to those names; and labels = "attributes" keeps the IDs and attaches readable labels (e.g. "Total > Male > Under 5 years") to the columns of wide output. load_variables(clean_labels = TRUE) returns the same cleaned labels.
  • load_variables() now includes each table's universe where the Census API publishes it (ACS detailed tables and the 2020 decennial files) (#596).
  • Requests to the Census API and Census Bureau file downloads now use {httr2} in place of {httr}. Rate limiting and temporary server errors are retried automatically, and connection failures and API outages give clear error messages.
  • Various bug fixes and performance improvements (#649, #590, #652, #557, #564, #582, #526, #651, #653, #654, #485).