Skip to content

Commit

Permalink
docs: clean up news for the release (#265)
Browse files Browse the repository at this point in the history
* docs: clean up news for the release

* prettify NEWS.md

* Update NEWS.md
  • Loading branch information
bms63 authored Mar 25, 2024
1 parent 3bf894b commit dbd5f67
Showing 1 changed file with 15 additions and 43 deletions.
58 changes: 15 additions & 43 deletions NEWS.md
Original file line number Diff line number Diff line change
@@ -1,76 +1,48 @@
# xportr (development version)

## New Features and Bug Fixes
## New Features

* All core functions can be run together by using new function `xportr()` (#137)
* `xportr_metadata()` can set `verbose` for a whole pipeline, i.e. setting `verbose` in `xportr_metadata()` will populate to all `xportr` functions. (#151)

* All `xportr` functions now have `verbose = NULL` as the default (#151)

* Bug fix for domain filtering (#137)

## Documentation

* Remove unused packages from Suggests (#221)

* `xportr_write()` now accepts `metadata` argument which can be used to set the dataset label to stay consistent with the other `xportr_*` functions. It is noteworthy that the dataset label set using the `xportr_df_label()` function will be retained during the `xportr_write()`.

* Exporting a new dataset `dataset_spec` that contains the Dataset Specification for ADSL. (#179)

* `xportr_split()` is a new function that allows users to split a dataset into multiple output files based on a variable. (#183)
* `xportr_write()` now accepts `metadata` argument which can be used to set the dataset label to stay consistent with the other `xportr_*` functions. It is noteworthy that the dataset label set using the `xportr_df_label()` function will be retained during the `xportr_write()`. (#179)
* Added a check for character variable lengths up to 200 bytes in `xpt_validate()`(#91, #189).

* File name check is moved to strict_checks condition to allow underscores in the file name. Underscores are allowed in xpt but not per FDA requirements. (#126)

* It is now possible to get and set the xportr options using the helper function `xportr_options()` (#130)

* Added `xportr.character_metadata_types` and `xportr.numeric_metadata_types` to list the metadata types that are character or numeric. Updated `xportr.character_types` and `xportr.numeric_types` to list only the R types that are character and the R types that are numeric. This ensures that all R types, including dates, are now managed by xportr_type. If the R type differs from the metadata type, the variable is coerced (#161)..

* Adds argument assertions to public functions using `{checkmate}` (#175)
* `xportr_split()` is a new function that allows users to split a dataset into multiple output files based on a variable. (#183)

* `xportr_metadata()` can set `verbose` for a whole pipeline, i.e. setting `verbose` in `xportr_metadata()` will populate to all `xportr` functions. (#151)

* All `xportr` functions now have `verbose = NULL` as the default. If left `NULL`, the previous default of `getOption("xportr.[fn_name]_verbose")` is used (#151)

* All core functions can be run together by using new function `xportr()` (#137)

* Added `xportr.character_metadata_types` and `xportr.numeric_metadata_types` to list the metadata types that are character or numeric. Updated `xportr.character_types` and `xportr.numeric_types` to list only the R types that are character and the R types that are numeric. This ensures that all R types, including dates, are now managed by xportr_type. If the R type differs from the metadata type, the variable is coerced (#161).
* New argument in `xportr_length()` allows selection between the length from metadata, as previously done, or from the calculated maximum length per variable when `length_source` is set to “data” (#91)

* Series of basic checks added to the `xportr_format()` function to ensure format lengths, prefixes are accurate for the variable type. Also to ensure that any numeric date/datetime/time variables have a format. (#164)

* Make `xportr_type()` drop factor levels when coercing variables

* `xportr_length()` assigns the maximum length value instead of 200 for a character variable when the length is missing in the metadata (#207)

## Bug Fixes

* Bug fix for domain filtering (#137)
* Make `xportr_type()` drop factor levels when coercing variables (#159)

## Deprecation and Breaking Changes

* The `domain` argument for xportr functions will no longer be dynamically
determined by the name of the data frame passed as the .df argument. This was
determined by the name of the data frame passed as the `.df` argument. This was
done to make the use of xportr functions more explicit. (#182)

* The `label` argument from the `xportr_write()` function is deprecated in favor of the `metadata` argument. (#179)
* The `metacore` argument, which was renamed to `metadata` in the following six xportr functions: (`xportr_df_label()`, `xportr_format()`, `xportr_label()`, `xportr_length()`, `xportr_order()`, and `xportr_type()`) in version `0.3.0` with a soft deprecation warning, has now been hard deprecated. Please update your code to use the new `metadata` argument in place of `metacore`.

* The `metacore` argument, which was renamed to `metadata` in the following six xportr functions: (`xportr_df_label()`, `xportr_format()`, `xportr_label()`, `xportr_length()`, `xportr_order()`, and `xportr_type()`) in version `0.3.0` with a soft deprecation warning, has now been hard deprecated. Please update your code to use the new `metadata` argument in place of `metacore`. (#203)
* `SASlength` and `SAStype` were removed since they did not have an impact on `xpt_validate` or any other functions (#132)
* Removes `admiral` from suggested dependencies (#237)
* `adsl` data object is now called `adsl_xportr` (#237)
* Data objects are no longer lazy loaded, which means that when needed the user must call `data("name_of_object")` first (#237)

## Documentation

* Created development version of the website (#187)

* Additional guidance for options added in deep dive vignette (#81)

* Added a vignette about standards in different agencies (#206)

* Removed non-user facing function documentation (#192)

* Added more details about `xportr_length()` and `xportr()` to deep dive vignette (#215, #222)

## Miscellaneous

* Tests use `{withr}` to create temporary files that are automatically deleted (#219)
* Remove unused packages from Suggests (#221, #237)
* Exporting a new dataset `dataset_spec` that contains the Dataset Specification for ADSL. (#179)
* Adds argument assertions to public functions using `{checkmate}` (#175)
* Data objects are no longer lazy loaded, which means that when needed the user must call `data("name_of_object")` first (#237)

# xportr 0.3.2

Expand Down

0 comments on commit dbd5f67

Please sign in to comment.