NEWS

xlr 1.1.2.9000

Added make_wider() function to easily pivot outputs from build_table() and related functions for creating wider table formats.
Added support for exclude_codes parameter in multiple response questions for the function build_mtable(). This parameter supports passing values like -99 that should be excluded from the analysis. See the documentation for build_mtable().
Added column width support to xlr_format(), instead of xlr trying to guess the width.
Created a new type xlr_n_percent, which combines a count and a percentage. Like all other xlr types, it can be written to Excel.

Added vector arithmetic support (vctrs::vec_math()) for xlr types. Most base R vector operations are now supported (e.g., sum(), mean(), median()). See vctrs::vec_math() for more details on supported operations.
xlr_integer now supports division and conversion to numeric (returns xlr_numeric).
Added support for type conversion between all xlr types. You can now convert all xlr types to each other, e.g., xlr_numeric to xlr_integer, xlr_percent to xlr_integer.
xlr_percent now works with ggplot2 without errors.
as.character() on xlr_percent now returns formatted percentage strings.
Refactored xlr classes to inherit from base types for better compatibility.

Fixed a bug in build_qtable() where an NA value in one column resulted in the entire row being removed.
Improved error handling in internal tidyr function calls.

Relaxed restrictions on column value requirements for build_qtable(). Response values for each column in the block only need to be a subset of the first column.
Improved documentation for question-related functions.
Added tests for ggplot2 compatibility.
Updated snapshots for remote tests.
Updated DESCRIPTION file for latest roxygen2 version.

CRAN re-submission:
- Fixed title and description to follow CRAN's standards (moved from back-ticks to single quotes)!