- Spatial boundaries will be cached using the DHS client (#122)
-
Convert DHS dataset flat file data dictionaries to UTF-8. This addresses parsing parsing issue that arose following R 4.3.0 (#146).
In doing so, it addresses issue of accents in variable and value labels not being parsed correctly.
- Update for deprecation of spatial infrastructure (#145).
- All geographic datasets are parsed using
sf::read_sf
(previouslyrgdal::readOGR
)
- Authentication patch for downloading datasets (#120)
download_boundaries
patch (#138) for pause when downloading survey boundaries. Set default sleep to 5 seconds to avoid timeout.
available_datasets
patch (#131) for new version ofqdapRegex
on CRAN
- Reference internal access to
model_datasets
withrdhs::model_datasets
to avoid errors ifrdhs
namespace is not loaded.
available_datasets
patch (#115)
-
Remove class
"dhs_dataset"
from downloaded micro data sets. This class is not anywhere and it creates an error for dplyr_v1.0.Cached datasets will need to be re-downloaded after updating to clear the dhs_dataset clas.
-
Replace
readLines()
withbrio::read_lines()
to make parsers robust to Windows encoding issues (similar to https://stackoverflow.com/questions/18789330/r-on-windows-character-encoding-hell). -
Use
"sf"
as default download method fordownload_boundaries(..., method = "sf")
. Add argumentsquiet_download
andquiet_parse = TRUE
todownload_boundaries()
.quiet_download
(defaultFALSE
) controlsdownload.file()
messages.quiet_parse
(defaultTRUE
) controls messages fromsf::st_read()
whenmethod = "sf"
.
- Add CITATION info.
- New
download_boundaries
for downloading spatial boundaries using (#71) - New
dhs_gps_data_format
for DHS GPS Information (#74) - Tibbles can be specified correctly as data.frame format (#89)
- Config creation on Windows 10 fixed (#91)
- Typos and messaging fixed (#78, #84, #87, #92)
unzip_special
correctly detects 4Gb files (#43)
- Addresses CRAN fail on windows
- New
delabel_df
for converting labelled data frames to characters (#54)
- Duplicate labels when parsing flat data files corrected (#79)
extraction(add_geo=TRUE)
correction for Kenya 2014 surveys (#67)- Geospatial covariate data sets now supported correctly (#64)
-
New
as_factor.labelled
for backward compatibility withhaven <2.0.0
labelled
classes. -
model_datasets
now internal and exported dataset (#60).
-
New vignettes:
country_codes
-
Documentation typos corrected (#55)
-
New
set_rdhs_config
for providing login credentials. This deprecatesset_dhs_credentials
. -
New
get_rdhs_config
shows the credentials currently used byrdhs
-
Permissionn from user now required for file saving.
-
API requests can now ignore any cached responses (
force = TRUE
) argument (#23):
dat <- dhs_countries(force = TRUE)
-
get_datasets(clear_cache = TRUE)
will clear the cached available datasets, enabling newly requested datasets to be downloaded (@kaisero, #29). -
geojson objects can now be requested from the API so that you can return geojson objects for mapping purposes (#28) e.g. :
d <- dhs_data(countryIds = "SN", surveyYearStart = 2014,
breakdown = "subnational", returnGeometry = TRUE,
f = "geojson")
# convert to spatial object
sp <- geojsonio::as.json(d) %>% geojsonio::geojson_sp
- New
dhs_data()
,dhs_countries()
and other API functions (dhs_x()
).
authenticate_dhs()
now works with short project names.
-
New vignettes:
anemia
-
New
read_dhs_flat()
for reading flat datasets.
- Initial share on Feb, 24th 2018 to colleagues at UNC.