R packages, especially the popular ones, are frequently updated. Although the packages go through a rigorous testing process before being accepted on CRAN, it is sometimes necessary to be able to save your work at a point in time. This comes up often during production development using R packages. Reproducibility of results are also oftentimes needed in regulated environments by regulators such as the FDA.
R provides some unique and mature capabilities to store and persist data at a point in time. To retrieve the results, all the user needs to do is simply revert to a pre-saved version of the work.
The three most popular methods of saving R work in a reproducible framework are:
- Packrat
- Checkpoint
- Rocker
