High-Performance Open-Source Archive
magrittr pipe removed: The
%>% pipe is no longer re-exported. Use the native R pipe
|> instead (available since R 4.1.0, which is already
the minimum version for rKolada). If you rely on %>% in
your own code, add library(magrittr) or switch to
|>.kolada_cache_dir(): New exported
function returning the persistent cache directory path (using
tools::R_user_dir()).kolada_clear_cache(): New exported
function for clearing cached files.kolada_available(): Now
performs a fast HTTP check (single request, 5s timeout) instead of
downloading actual data. Much faster in examples and tests.simplify = TRUE explanation, inline ggplot comments, and
“K”-filter explanations.glue replaced by cli: All
string interpolation now uses cli for user-facing messages
and paste0() for URL construction. glue is no
longer a dependency.magrittr removed: All internal code
uses the native pipe |>. magrittr is no
longer a dependency.message(),
warning(), and stop() replaced with
cli::cli_inform(), cli::cli_warn(), and
cli::cli_abort() throughout.[func()] instead of
\code{\link{func}}).kolada_get() internal helper:
Centralised HTTP request logic with tryCatch(), timeout,
and retry.CRAN-RELEASE file.test_api.R test stub.auspices column
returned by get_kpi() is now named auspice
(matching the v3 API schema).ou_publication_date removed: This
field is no longer returned by the v3 API and has been removed from
kpi_describe() output.is_divided_by_gender: Now returns a
logical (TRUE/FALSE) instead of integer
(0/1).remove_undocumented_columns removed:
kpi_minimize() no longer accepts the
remove_undocumented_columns parameter as the API has been
updated with proper documentation for all data properties.https://api.kolada.se/v3/ instead of
http://api.kolada.se/v2/. The v2 API is being shut down on
March 31, 2026.get_municipality() gains a region_type
parameter to filter by municipality type (e.g., "K" for
municipalities, "L" for regions).get_values() gains a from_date parameter
to fetch only data updated after a specific date.get_values() gains a keep_deleted
parameter (default FALSE). Rows marked as deleted in the v3
API are filtered out by default.next_url field and supports
up to 5000 results per page (up from 2000).kolada_available() is now more resilient to v3 response
shape changes.httr replaced by httr2:
The HTTP backend has been migrated from httr to
httr2. This is an internal change with no user-facing API
differences.urltools removed: No longer a
dependency.dplyr::filter_at() and
dplyr::select_if() replaced with modern
dplyr::if_any() / dplyr::all_of()
equivalents.entity_search() helper eliminates code
duplication across kpi_search(),
municipality_search(), and ou_search().append_query_params() helper replaces
urltools::param_set()..data$ pronoun usage in tidyselect contexts
replaced with string column names.get_values() and get_metadata() now
automatically chunk oversized path parameters into batches of 25 to
respect the Kolada v3 API limit. Previously, passing more than 25 KPIs,
municipalities, or periods would return an error. Chunking is
transparent — results are combined into a single tibble.kolada_available() for programming with the
rKolada packageverbose = TRUE (#6)kolada_available()max_results argument to all get_
functionsget_values()
Need mirroring services?
Contact our team at info@vpspulse.com.
Mirror powered by VPSpulse
Infrastructure sponsored by VPSPulse & Secure Payments by ArionPay.