R Package Testing FAQs

Question 1

What is the difference between test_that() and describe()?

Accepted Answer

Use test_that() for ensuring functional correctness (doing things right), while describe() and it() use BDD syntax to verify behavioral specifications (implementing the right things).

Question 2

How should I handle external API calls during R tests?

Accepted Answer

Use local_mocked_bindings() to replace external functions with mock implementations that return predictable data, allowing you to test logic without making actual network calls.

Question 3

What are snapshot tests in R?

Accepted Answer

Snapshot tests record the exact output of a function—such as a plot, a complex data frame, or an error message—and compare future runs against that saved version to detect unintended changes.

Question 4

How do I ensure my tests don't leave mess behind on my system?

Accepted Answer

Use the withr package (e.g., withr::local_tempfile() or withr::local_options()) to ensure that any changes to the environment or file system are automatically reverted when the test finishes.

Question 5

How do I upgrade my R package to use testthat 3rd edition?

Accepted Answer

You can initialize or upgrade your package by running usethis::use_testthat(3) in your R console, which updates your DESCRIPTION file and sets the testing edition.

R Package Testing

Key Features

Use Cases

R Package Testing

Key Features

Use Cases