worldbank 0.10.0
- The PIP functions now require
release_version and
ppp_version to be exactly 8 and 4 digits, given as a string
or a number.
fone_dataset() and fone_view() now return
no more than the requested limit when pagination is
required.
fone_dataset(), fone_view(), and the PIP
functions now keep Namibia’s "NA" country code, instead of
reading it as a missing value.
fone_view() now trims whitespace and returns
NA instead of "" for empty fields, matching
fone_dataset() and the rest of the package.
pip_cp(), pip_data(),
pip_group(), and pip_versions() now trim
whitespace and return NA instead of "" for
empty fields, matching the rest of the package.
pip_cp(), pip_data(), and
pip_group() now require povline to be a number
between 0 and 2700, and
pip_data() and pip_group() require
popshare to be a number between 0 and
1.
pip_data() and pip_group() now correctly
document popshare as taking precedence over
povline; the documentation previously claimed the opposite.
The behavior is unchanged.
wb_bulk() now keeps Namibia’s "NA" code in
the iso2_code and wb_iso2_code columns of the
country table, instead of reading it as a missing
value.
wb_data() gains a footnote argument to
return the footnotes published with each observation.
wb_data() no longer returns NA dates when
a query mixes annual and sub-annual indicators.
wb_project() now returns results when
country holds more than one code, instead of silently
returning no rows. Codes are also upper-cased before querying, and must
be two characters, since the Projects API matches them exactly.
wb_project() now accepts multiple values for
id, status, and region, returning
the projects that match any of them.
wb_project() now requires status to be one
of "active", "closed", "dropped",
or "pipeline", matched case insensitively.
wb_project() now numbers the rows of the returned data
frame, instead of naming them after the project IDs already held in the
id column.
wb_search() now numbers the rows of the returned data
frame, instead of naming them after each match’s position in the
indicator catalog.
worldbank 0.9.1
wb_data() now returns an empty data frame instead of
erroring when a query has no observations.
worldbank 0.9.0
wb_bulk() downloads the entire World Development
Indicators dataset as a zip and returns its six CSVs as a list of data
frames, including footnote and series-time metadata not exposed by the
API.
wb_search() searches the indicator catalog for a
regular expression pattern across one or more fields.
- Documentation improvements.
worldbank 0.8.0
wb_data() gains mrv and
gapfill parameters to query the most recent values and
forward-fill missing data.
wb_project() queries World Bank project data with
support for filtering by country, status, region, date range, and
free-text search.
pip_cp() returns country profile data from the PIP
API.
pip_data() gains a nowcast parameter to
include nowcast estimates.
wb_cache_clear() replaces
wb_cache_delete(), which is now deprecated.
worldbank 0.7.1
- Don’t cache queries in tests to comply with CRAN
worldbank 0.7.0
- Added optional request caching to improve performance for repeated
queries. Enable with
options(worldbank.cache = TRUE). Cache
duration can be customized with worldbank.cache_max_age
(default: 1 day).
- New functions
wb_cache_delete() and
wb_cache_dir() to manage the cache.
worldbank 0.6.1
- Documentation improvements.
worldbank 0.6.0
wb_data() has been added as an alias for
wb_country_indicator()
worldbank 0.5.0
wb_country_indicator() now allows for multiple
indicators to be requested at once.
worldbank 0.4.0
- Removed conditional return of tibble and now always return a
data.frame instead. This ensures type stability across all
functions.
- Add support for Finances One API data.
worldbank 0.3.0
- Moved to HTTPS for all API calls.
- Support for Poverty and Inequality Platform (PIP) data.
worldbank 0.2.0
- Support month and quarter date format in
wb_country_indicator().
- More robust error handling.
worldbank 0.1.0