- Rules specified under the
all_datasets
keyword in a format list will apply to every data set of the reformatted object unless specified otherwise. - New
verbose
argument in thereformat
method. When applied tolist
the value of this augment can be controlled with thedunlin.reformat.verbose
option or theR_DUNLIN_REFORMAT_VERBOSE
environment variable. - Improve the output when printing
rule
objects. - New
combine_rules
andcombine_list_rules
functions to combine rules or list of rules into a single rule or a single list of rules.
render_safe
now renders placeholder using in priority values corresponding to the key matching exactly the placeholder, case included.- New
show_whisker
function to display the available whiskers. join_adsub_adsl
now sends a warning when the continuous or categorical variables are missing or allNA
. AllNA
variables can now be kept usingdrop_na = FALSE
. In addition, missing levels in the pivoted columns can be dropped withdrop_lvl = TRUE
.- Remove
mini_pivot_wider
function which is no longer used.
empty_rule
is removed now.rule()
will create a normalrule
object.- By default
rule
are converting empty space""
toNA
upon reformatting. list2rule
now tolerates duplicated rules with different names.- Specified minimal version of package dependencies.
rules
now have attributes controlling their behavior during reformatting.reformat
can overriderule
attribute with the corresponding arguments.- Add
render_safe
andadd_whisker
functions to replaces placeholders enclosed in curly braces with replacement stored in a dedicated environment. read_rule
function has been moved into thecitril
package.
- Add
log_filter
function for filtering with additional logs. - Remove the use of
dm
. - Introduced
ls_explicit_na
to replaceNA
values in an entirelist
ofdata.frame
. reformat
re-levels factors to follow the order provided in the rule that is applied and levels declared in the rule are all integrated to the result.
- Defined a Reformatting Map (
rule
object) which specifies the correspondence between the old and the new values in data. - Enhanced function
dunlin::reformat()
that provides generic supports incharacter
and/orfactor
reformatting while keeping the attributes of the original data (e.g. the data type or labels) unchanged. - Improved performance of package by adding more unit tests.
- Upgraded package dependencies and removed deprecated functions.
- Updated
pkgdown
configuration.
- Initial package release.