Introducing job::job()

The job package is on CRAN and judging from the absence of bug-reports, it seems to be doing well! It runs code as an RStudio job, so your console is free in the meantime. It’s very useful to stay productive while doing cross-validation, bayesian MCMC-based inference, etc.

The default usage is simply to wrap your code:

df = mtcars[mtcars$gear == 3]
job::job({
  library(brms)
  fit = brm(mpg ~ wt + (1|cyl), df)
})

… and then it returns the results to your main session when it’s done (fit, in this case). You can launch multiple jobs and keep track of them in RStudio’s “jobs” pane:

Check out the job website to read more use cases and the many options for customizing the call.