toBaseR() gains a method for data.table: key, secondary indices and
the internal self-reference are dropped, and a deep copy is returned.
haven_labelled columns are converted to factors, as for tibbles.toBaseR() on tibbles no longer breaks up list and matrix columns.courseData() has been renamed to readCourseData(), which describes the
function better.resolveFormula(): subset now follows base R semantics (evaluated in
data, as in lm()); new resolveFormulaFromCall() for formula methods.
y ~ a:b is accepted as the cells of several grouping variables.conceptMap() and conceptAudit(); getConcepts() is now exported.Parts of the code and documentation were reviewed with the help of large language models (OpenAI Codex, Anthropic Claude). Every suggestion was assessed, edited and verified by the maintainer, who remains solely responsible for the content of this package.