R Markdown — Dynamic Documents and Reports for R

Introduction

R Markdown is an authoring framework from Posit (formerly RStudio) that lets analysts write documents combining prose, code, and results in a single file. When rendered, code chunks execute and their output (tables, plots, values) is woven into the final document. It is the standard approach for reproducible reporting in the R community.

What R Markdown Does

• Embeds executable R, Python, SQL, and Bash code chunks inside Markdown narrative
• Renders to HTML, PDF (via LaTeX), Word, PowerPoint, and many other formats through Pandoc
• Supports parameterized reports where input values change the output without editing the source
• Provides output templates for articles, slides (ioslides, Slidy, reveal.js, xaringan), dashboards (flexdashboard), and websites
• Integrates with knitr for code execution and caching, and with Pandoc for format conversion

Architecture Overview

An R Markdown file (.Rmd) is processed in two stages. First, knitr reads the file, executes each code chunk in an R session, captures the output, and produces a plain Markdown (.md) file. Second, Pandoc converts the Markdown to the target format (HTML, PDF, DOCX). R Markdown's YAML front matter specifies the output format and options. Custom output formats can be defined as R functions that configure Pandoc flags and provide templates.

Self-Hosting & Configuration

• Install the rmarkdown package from CRAN; Pandoc is bundled with RStudio or can be installed separately
• Configure output options in the YAML header: title, author, date, output format, and format-specific settings
• Use params: in the YAML header for parameterized reports that accept input values at render time
• Enable code chunk caching with knitr::opts_chunk$set(cache = TRUE) to speed up iterative rendering
• Deploy rendered HTML reports to RPubs, RStudio Connect, or any static web host

Key Features

• Single-source reproducibility: the document IS the analysis, ensuring results always match the code
• Parameterized reports let users generate customized outputs without modifying the source
• Chunk options control execution, display, caching, figure size, and error handling per chunk
• Inline R expressions (r expr) embed computed values directly in narrative text
• Ecosystem of extension packages adds specialized formats: flexdashboard, xaringan, distill, blogdown

Comparison with Similar Tools

• Quarto — the next-generation successor to R Markdown, supporting Python, Julia, and Observable alongside R; R Markdown remains stable and widely used
• Jupyter Notebook — supports many languages and is dominant in Python; R Markdown produces cleaner version-controlled source files
• Marimo — reactive Python notebooks; R Markdown is batch-rendered and document-oriented rather than interactive
• LaTeX — full typesetting power; R Markdown provides a simpler Markdown front-end that can still output through LaTeX
• Org Mode (Emacs) — literate programming in Emacs; R Markdown is more accessible and has broader tool support

FAQ

Q: What is the difference between R Markdown and Quarto? A: Quarto is the successor with multi-language support and a standalone CLI. R Markdown is R-centric and relies on the rmarkdown R package. Both produce similar output.

Q: Can I use Python code in R Markdown? A: Yes. Use {python} code chunks. The reticulate package bridges R and Python, allowing data sharing between the two.

Q: How do I create a PDF from R Markdown? A: Set output: pdf_document in the YAML header. A LaTeX distribution (TinyTeX recommended) must be installed for PDF rendering.

Q: Is R Markdown suitable for academic papers? A: Yes. The rticles package provides templates for many journals. R Markdown handles citations, cross-references, and bibliographies.

Sources

• https://github.com/rstudio/rmarkdown
• https://rmarkdown.rstudio.com/