How to Get Help With R Packages

Q: How do I see all help topics in a package?

Use help(package = "packageName") to open the package help index. You can also use browseVignettes("packageName") to find its available tutorials.

R packages add functions, data, and other tools to R. When a function is unfamiliar or an error interrupts your work, there are several ways to find reliable help. Begin with the package’s own documentation, which is often the fastest route; if that does not answer your question, search or ask in a community forum with enough detail for someone to reproduce the issue.

What You’ll Need

An R session, such as R in a terminal or RStudio.
The package name and, if relevant, the function name.
The error message or a short description of what you expected and what happened.
A small sample of the data and code that demonstrates the problem, with private or sensitive information removed.

Step-by-Step

Check that the package is available. If the package is already installed, load it with library(packageName), replacing packageName with its actual name. If R says there is no package by that name, check the spelling and whether it is installed. For a package hosted on CRAN, install it with install.packages("packageName"), then load it. Installation and loading are separate actions.
Open the function’s help page. Use ?functionName or help("functionName"). For example, ?read.csv opens help for a function. If the name is ambiguous or you want to specify a package, use help("functionName", package = "packageName"). The help page describes arguments, return values, details, and examples.
Look at the examples. Find the Examples section in the help page, or run example("functionName"). Try a simple example before applying the function to your full project. Examples can reveal required input formats and show how arguments work together.
Read the package vignette or official guide. A vignette is a longer tutorial that explains a package workflow, often with runnable code. Use browseVignettes("packageName") to find available vignettes in an installed package. You can also check the package’s CRAN page or its official documentation site. Confirm that the guide applies to your installed package version.
Search the exact problem. Copy a distinctive part of the error message into a search, together with the package and function names. Search results may include useful discussions on Stack Overflow, Posit Community, or the package’s issue tracker. Read the accepted or maintainer-provided answer carefully, and check whether it matches your R and package versions.
Make a minimal reproducible example. Remove unrelated code and data until the issue still occurs in a short example. Use a small sample rather than sharing a large, private, or confidential dataset. For data-frame examples, dput(head(my_data)) can provide a reproducible representation of a few rows; review the output first to ensure it contains no sensitive information. Include the code, error text, and what you expected to happen.
Ask in the right place. For a general R question, use a community forum such as Stack Overflow or Posit Community and follow its posting rules. For a suspected package bug, first check the package’s issue tracker for an existing report; use its issue template if you open a new one. Do not post the same question in several places at once. Be polite, specific, and willing to clarify details.
Follow up with what you tried. Say which help pages or examples you checked, what changed when you tried a suggested fix, and whether the issue is resolved. If the answer depends on your setup, include relevant details from R.version.string, packageVersion("packageName"), and sessionInfo(). Share only the output needed to diagnose the problem.

Common Mistakes

Confusing installation with loading: install.packages() adds a package to your library; library() attaches an installed package to the current session. Installing again usually does not solve a problem caused by a missing library() call.
Leaving out the exact error: A summary such as “it does not work” hides important clues. Copy the full error and, when relevant, the warning that appeared immediately before it.
Sharing too much or too little code: A full project is difficult to inspect, but a fragment that cannot reproduce the issue is not enough. Keep only the smallest code and data that still show the problem.
Assuming every online answer applies: An answer may refer to another package version, operating system, or function. Check the context and verify that the proposed solution fits your situation before using it.
Posting private data or credentials: Remove names, identifying information, passwords, access tokens, and proprietary material before sharing code or output.

Tips

Search the package and function names along with a short, distinctive phrase from the error. Quoting an exact error fragment can make results more relevant.
Use ??keyword to search installed help topics when you do not know the exact function name.
Use args(functionName) for a quick look at a function’s arguments; use its full help page for explanations and behavior.
Use packageVersion("packageName") to check which version is installed. Include this when reporting behavior that may depend on a release.
When a function name exists in more than one package, call it explicitly as packageName::functionName or search its help with the package specified.
For package-specific questions, look for contribution guidelines or a support section before opening an issue. Issue trackers are generally best for reproducible bugs, not general R lessons.

Troubleshooting

“There is no package called…”: Check the package spelling, install it if appropriate, and make sure installation completed without an error. If it is not on CRAN, consult the package’s official installation instructions rather than guessing an installation command.

“Could not find function…”: Confirm that the package is installed and loaded. The function may have a different name, may belong to another package, or may not be exported for direct use. Check the package help index or documentation.

The help page does not appear: Verify the function spelling and whether the package is installed. Try help("functionName", package = "packageName"); for a web page, check that your browser can open the package documentation site.

Online solutions do not match your result: Compare the R and package versions shown in the solution with yours. Re-run the smallest example in a fresh R session if practical, and mention any relevant differences when asking for help.

You cannot reproduce the issue with a small example: Add back the parts of the original code gradually until the failure returns. This can identify a particular input, option, or package interaction. Keep the example safe to share.

Frequently Asked Questions

How do I see all help topics in a package?

Use help(package = "packageName") to open the package help index. You can also use browseVignettes("packageName") to find its available tutorials.

What is a reproducible example in R?

It is a compact set of code and sample data that lets another person run the same steps and see the same problem. It should not depend on files or private information they cannot access.

Should I ask on Stack Overflow or open a package issue?

Ask a community for help using R or understanding a function. Use the package’s issue tracker for a suspected bug or a feature request, after checking existing issues and following the project’s guidelines.

What information should I include when asking for help?

Include the package and function names, relevant versions, a minimal example, the full error message, and what you expected. Remove confidential data and credentials before posting.

Can I get help directly in RStudio?

Yes. You can run R help commands in its console, and RStudio commonly provides a Help pane for help topics. The interface can differ by version, but the R commands work independently of that pane.

Further reading: R Project. Details can change, so check the official page for anything that matters to you. Spotted a mistake? Tell us. How we write an answer.