Changes in version 0.2.0.9003                      

  - Development version: Fix handling of variable extraction for
    DBIsvydesign objects. Add several warnings, errors, documentation
    notes, and test cases.

                     Changes in version 0.2.0.9002                      

  - Development version: Thanks to Ben Schneider @bschneidr for all of
    the following updates.
  - Robustify folds.svydesign() and cv.svydesign() to extract
    information from svydesign objects more safely, enabling them to
    work with srvyr package as well as designs that have been subsetted
    (e.g. with subset()) or updated (e.g. with update() or transform()).
  - Throw informative warnings (if trying to use a multi-stage sampling
    design) and errors (if trying to use a replicate design or not
    providing a design object).
  - Allow cv.svydesign() to extract information from a DBIsvydesign
    object. (Currently, just turns it into a regular svydesign object
    after extracting the minimal necessary data.frame. Dealing with
    DBIsvydesign objects natively is on the to-do list.)

                 Changes in version 0.2.0 (2022-03-15)                  

  - Make folds.svy() and folds.svydesign() functions directly accessible
    and documented, so that users can form survey-design-based CV folds
    and use them with other models besides svyglm.
  - Add examples to README and intro vignette illustrating how to use
    folds.svy() with design-consistent random forest models from the
    rpms package.

                        Changes in version 0.1.2                        

  - Allow logistic regression response variable to be a factor, not
    just 0/1. (Thanks to Aja Sutton @amsutton and Michelle Jamieson
    @themichjam.)
  - Add logistic regression example() to function documentation.
  - Add na.rm arguments, passed internally to svymean when calculating
    average of CV test losses.

                     Changes in version 0.1.1.9001                      

  - Development version: update docs with link to Stat
    doi:10.1002/sta4.454 following online publication on Jan 12, 2022.

                     Changes in version 0.1.1.9000                      

  - Development version: update docs with CRAN link and installation
    instruction, and with news about the paper's acceptance to Stat on
    Jan 9, 2022.

                 Changes in version 0.1.1 (2022-01-10)                  

  - Initial package release, as accepted to CRAN on Jan 10, 2022.

                        Changes in version 0.1.0                        

  - Initial package pre-release. Includes old simulation results
    reported in early-results vignette based on code in
    R/plot_generation.R, to be removed before CRAN submission.