Predict first
Before you run the code, decide what result would make sense. For the binomial example, try calculating the probability by hand first.
An actual statistical computing environment
Real R. In your browser.
Use the language statisticians actually use. Change a distribution, fit a model, reproduce a published comparison, or write your own code. Your calculations run locally through WebAssembly.
Optional computing laboratory: No R code runs until you press Run R. The browser downloads the webR runtime from its official host on first use, which may take a moment. Your code runs in your browser, not on our server. You can also copy the code into desktop R or RStudio.
Statistics · R Workbench · Beta
Change the example. Keep the evidence.
Explore with base R. Choose a prepared lesson example or edit any part of the script. The selected example never runs automatically.
Mean, median, variance, and a histogram · Read the related lesson →
R has not started. Select an example, then press Run R.
R console
Output appears here after you run the script.
R graphics
No plots yet.
Base R plotting is available, but this is not a managed compute service. A long-running or infinite loop may require reloading the tab, especially with the PostMessage runtime channel. The workbench has no external package installation button and no automatic server-side data storage.
Try a real dataset: The historical New York air-quality field notebook explains units, provenance, missing readings, and why the regression's sample changes before the model runs. Choose “Audit missing measurements” or “Model real New York air quality” in the workbench.
How to work with it
Change one thing
Replace a sample, alter a seed, or change a model. Inspect what happened rather than simply trusting a p-value.
Explain what survived
Ask which statistical assumptions your change preserved, which it violated, and whether you could reproduce the result elsewhere.
Runtime implementation: webR. For more examples, explore the interactive laboratories and research bibliography.