Contributing to lsastrat

Thanks for your interest in improving lsastrat. Contributions of all kinds are welcome: bug reports, documentation fixes, methodological discussion, and code.

Reporting a bug or asking a question

• Search the issue tracker first.
• For a bug, please open an issue with a minimal reprex (reprex package) — a small, self-contained example (the bundled pisa_mini data set is ideal) that reproduces the problem, plus the output of sessionInfo().
• For usage questions or methodological discussion, open a Discussion or a question issue. Because the package implements specific survey-statistics methods, please point to the relevant formula or reference where you can.

Contributing code

1. Fork the repository and create a feature branch off main.

2. Follow the existing style (roughly the tidyverse style guide: two-space indentation, snake_case, <- assignment). Keep functions small and documented with roxygen2.

3. Add or update testthat tests for any change in behaviour. New estimators or variance code should include a test that recovers a known value or cross-validates against an established package where possible.

4. Run the checks locally before opening a pull request:

   devtools::document()   # regenerate man/ and NAMESPACE
   devtools::test()       # run the test suite
   devtools::check()      # R CMD check

5. Update NEWS.md with a bullet describing user-facing changes.

6. Open a pull request describing the change and the motivation. Link any related issue.

Methods changes

lsastrat aims to be methodologically defensible. If you change how a quantity is estimated (pooling, variance, bias correction, a new index), please include in the pull request: the reference it follows, and either a simulation showing the estimator recovers the target or a cross-check against another implementation.

Code of conduct

Please note that this project is released with a Contributor Code of Conduct. By participating you agree to abide by its terms.