New Feature: mask_names() gains a keep_suffixes parameter to preserve a fixed suffix (e.g. _r for reverse-scored items) verbatim in masked names, instead of masking it away. When multiple supplied suffixes match the same column, the longest match is kept and a warning reports the affected column(s).
Behavior Change: mask_names() now sorts masked columns alphabetically by their masked name within their original positions, so column position no longer leaks which original variable a masked name refers to. Unmasked columns keep their original positions.
Bug Fix: scramble_variables() now selects target columns by name instead of by numeric index, fixing a bug where a .groups column positioned before or between target columns could cause the wrong column to be scrambled.
Validation: mask_names(), mask_variables(), and scramble_variables() now reject data frames with duplicate column names with a clear error, instead of relying on inconsistent errors from underlying dependencies.
Documentation: Corrected the source citations for the williams and marp example datasets, and removed a placeholder OSF link.
Update documentation and vignettes to reflect the new keep_suffixes parameter.
vazul 1.1.0
2026-02-06
Breaking Change: Removed mask_variables_rowwise() as row-wise masking is methodologically inconsistent with analysis blinding.
Breaking Change: scramble_variables_rowwise() is now internal. Its functionality has been consolidated into scramble_variables(..., .byrow = TRUE).
API Improvement: Arguments in scramble_variables() and mask_variables() parameters now use leading dots (e.g., .groups, .together) to avoid naming conflicts with data columns.
Update documentation, vignettes, and website to reflect changes in function names and parameters.
vazul 1.0.0
2026-01-16
Finalize functions
Streamline validation and generics
Unify the API interface for functions
Add tests, change vignettes to match functionality
