Home

Awesome

targets package minimal example

Launch RStudio Cloud

This repository is an example data analysis workflow with targets. The pipeline reads the data from a file, preprocesses it, visualizes it, and fits a regression model.

How to access

You can try out this example project as long as you have a browser and an internet connection. Click here to navigate your browser to an RStudio Cloud instance. Alternatively, you can clone or download this code repository and install the R packages listed here.

How to run

  1. Open the R console and call renv::restore() to install the required R packages.
  2. call the tar_make() function to run the pipeline.
  3. Then, call tar_read(hist) to retrieve the histogram.
  4. Experiment with other functions such as tar_visnetwork() to learn how they work.

File structure

The most important files are:

├── _targets.R
├── R/
├──── functions.R
├── data/
├──── raw_data.csv
└── index.Rmd
FilePurpose
_targets.RThe special R script that declares the targets pipeline. See tar_script() for details.
R/functions.RAn R script with user-defined functions. Unlike _targets.R, there is nothing special about the name or location of this script. In fact, for larger projects, it is good practice to partition functions into multiple files.
data/raw_data.csvThe raw airquality dataset.

index.Rmd: an R Markdown report that reruns in the pipeline whenever the histogram of ozone changes (details).

Continuous deployment

Minimal pipelines with low resource requirements are appropriate for continuous deployment. For example, when this particular GitHub repository is updated, its targets pipeline runs in a GitHub Actions workflow. The workflow pushes the results to the targets-runs branch, and GitHub Pages hosts the latest version of the rendered R Markdown report at https://wlandau.github.io/targets-minimal/. Subsequent runs restore the output files from the previous run so that up-to-date targets do not rebuild. Follow these steps to set up continuous deployment for your own minimal pipeline:

  1. Ensure your project stays within the storage and compute limitations of GitHub (i.e. your pipeline is minimal). For storage, you may choose the AWS-backed storage formats (e.g. tar_target(..., format = "aws_qs")) for large outputs to reduce the burden on GitHub storage.
  2. Ensure GitHub Actions are enabled in the Settings tab of your GitHub repository’s website.
  3. Set up your project with renv (details here).
    • Call targets::tar_renv(extras = character(0)) to write a _packages.R file to expose hidden dependencies.
    • Call renv::init() to initialize the renv lockfile renv.lock or renv::snapshot() to update it.
    • Commit renv.lock to your Git repository.
  4. Write the .github/workflows/targets.yaml workflow file using targets::tar_github_actions() and commit this file to Git.
  5. Push to GitHub. A GitHub Actions workflow should run the pipeline and upload the results to the targets-runs branch of your repository. Subsequent runs should add new commits but not necessarily rerun targets.