fast (see one example of timings below).

reproducible workflows – rerun-tolerant.

Install and Require are identical except that Require will also call require (lower case r) on all the named packages with the default setting of require = TRUE. “rerun-tolerant” is a requirement for a robust reproducible workflow for every “manual” break in code (i.e., a user runs a bit of code, then skips a few lines, then runs more etc.) provides the potential for sections of code to become stale without the user being aware. This function uses RAM caching, so the first time it is run in a new R session will be slower than subsequent times in which cached copies of e.g., the package dependency tree, can be used. If no packages are missing, then it will not install anything.

This means that if one or more packages is not installed prior to running the function, then the function will determine which are not installed, install those. “rerun-tolerant” means that the results from running this function (the output) will be identical each time, even when the conditions when run are different. The Require package provides two “rerun-tolerant” functions, Require and Install (a recent addition). For example, what is the best way to move analyses from one machine to another, or set up a series of High Performance Compute nodes? How should we use functions like install.packages in a reproducible workflow that are clearly intended to be used once or very few times? How do we deal with many packages on GitHub that have many common dependencies? How do we deal with packages that have dependencies that are no longer on CRAN (“they have been archived”)? How do we replicate an analysis 6 months from now when some packages have changed, and their dependencies have changed? Finally, how do we do all this for many concurrent projects without installing hundreds of packages in a new directory for every project?

Mixing many package dependencies that are constantly evolving creates challenges with standard R package management. Some of them are on CRAN, but some are still in development and so are hosted elsewhere. Some packages, including those in our PredictiveEcology repository, have many package dependencies.

R update packages