Contributing to xpose

We are happy that you are willing to help us out with the development of xpose. Before get started please take some time to read the following guidelines.

Submitting issues

Before posting a new issue, please check that it has not already been posted.

  • For feature requests make sure to include sufficient information, such as a proposed workflow, plots images, links to posters or publications describing a method, etc.

  • For reporting a bug, the most important thing is to include a minimal reproducible example so that we can quickly verify your problem, and figure out how to fix it. To make your example reproducible:

    • Attach files such as the NONMEM output files or an xpdb saved as .Rdata (i.e. save(xpdb, "xpdb_error.Rdata")). Note, you can use the arguments ignore and/or extra_files of xpose_data() to reduce the size of the xpdb whenever needed.
    • Include the R code that created the error. Ensure that your code is clear, well commented and as small as necessary to reproduce the error.
    • Include the error message.
    • Include information on your R session using sessionInfo().

Submitting pull requests

To contribute code to xpose please carefully follow the steps listed below.

Note 1: before getting started with a pull request, we strongly encourage you follow the procedure described above for feature requests and/or reporting a bug to ensure that your project is compatible with the our xpose road-map.

Note 2: if you are unfamiliar with the development of R packages, please take some time to read the R packages book by Hadley Wickham.

  1. Fork the xpose repository.

  2. Make the changes to the code on your fork.

    • Use the tidyverse coding style.
    • Any new function should have Roxygen documentation. Internal functions should contain #' @keywords internal. Add examples to new functions. Make sure to update the documentation with devtools::document() before submitting the pull request.
    • Tests should be added under tests/testthat/ to ensure for proper code coverage.
    • New plots should be created using the generic xplot_scatter(), xplot_distrib() or xplot_qq(). Check for example the code of R/plot_gof.R to see how it should be done.
    • General utility functions should be added under R/utils.R and plot utility functions under R/xplot_helpers.R.
    • Use functions from the tidyverse (e.g. dplyr, tidyr, stringr, etc.) whenever possible.
    • Add a brief description of your contribution to NEWS.md using bullets. Make sure to list it under the most recent xpose version and to add your name as (@<your_github_username>). e.g. * Added support for multiple pages plots (@guiastrennec). In addition, if your pull request addresses one or several issues, please reference them using the # sign e.g. * Added support for multiple pages plots (@guiastrennec, #3)

  3. Run devtools::check() on your fork and ensure the absence of any errors, warnings and notes.

  4. Create a pull request to the dev branch of the UUPharmacometrics/xpose repository.

    • Provide a good description of your pull request. Add examples, function outputs and plots images illustrating the new features.
    • Make sure that travis-ci, appveyor and codecov tests are all sucessfull

  5. Address all comments until the pull request is either merged or closed.

All of this may seem complicated at first, but you can contact one of the core developers to assist you throughout the process.