NEWS

---

dunlin 0.1.8.9002

dunlin 0.1.8

• 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 the reformat method. When applied to list the value of this augment can be controlled with the dunlin.reformat.verbose option or the R_DUNLIN_REFORMAT_VERBOSE environment variable.
• Improve the output when printing rule objects.
• New combine_rules and combine_list_rules functions to combine rules or list of rules into a single rule or a single list of rules.

dunlin 0.1.6

• 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 all NA. All NA variables can now be kept using drop_na = FALSE. In addition, missing levels in the pivoted columns can be dropped with drop_lvl = TRUE.
• Remove mini_pivot_wider function which is no longer used.

dunlin 0.1.5

• empty_rule is removed now. rule() will create a normal rule object.
• By default rule are converting empty space "" to NA upon reformatting.
• list2rule now tolerates duplicated rules with different names.
• Specified minimal version of package dependencies.

dunlin 0.1.4

• rules now have attributes controlling their behavior during reformatting.
• reformat can override rule attribute with the corresponding arguments.
• Add render_safe and add_whisker functions to replaces placeholders enclosed in curly braces with replacement stored in a dedicated environment.
• read_rule function has been moved into the citril package.

dunlin 0.1.3

• Add log_filter function for filtering with additional logs.
• Remove the use of dm.
• Introduced ls_explicit_na to replace NA values in an entire list of data.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.

dunlin 0.1.2

• 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 in character and/or factor 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.

dunlin 0.1.1

• Initial package release.