• allow duplicated clean values and labels so raw values can be collapsed
  • increment for change in contact information

Bug fix

  • Bug fix for #3 (h/t @ekatef), in which columns with uppercase names were dropped by default due to internal conflict with case_ignore and drop_extra options.

Bug fix

  • Bug fix for issue #2 due to update of labelled package to 2.1.0
  • encoded vector now output in same class as clean values from crosswalk file; this means that if the raw variable is a character (like state name) and the clean encoding value is an integer (like FIPS code), then the new labelled vector will be an integer class. This is the logical behavior and should work better with transfer to Stata, which only takes labels on numeric classes.
  • new tests
  • adjustments to account for new haven labelled classes
  • various improvements and bug fixes
  • update documentation
  • update vignette
  • Initial CRAN release