Literate data analysis with Stata and Markdown

In this article, I introduce markstat, a command for combining Stata code and output with comments and annotations written in Markdown into a beautiful webpage or PDF file, thus encouraging literate programming and reproducible research. The command tangles the input separating Stata and Markdown code, runs the Stata code, relies on Pandoc to process the Markdown code, and then weaves the outputs into a single file. HTML documents may include inline and display math using MathJax. Generating PDF output requires access to LATEX and a style file from Stata but works with the same input file.本文介绍了markstat命令，该命令可将Stata代码、输出结果与采用Markdown编写的注释及批注相结合，生成美观的网页或PDF文件，从而推动文学编程（literate programming）与可复现研究（reproducible research）的发展。该命令会对输入内容进行解缠处理，分离出Stata与Markdown代码，运行Stata代码后，依托Pandoc处理Markdown代码，最终将各类输出整合为单个文件。HTML文档可通过MathJax支持行内公式与块级公式的渲染。生成PDF输出则需使用LaTeX及Stata提供的样式文件，但可复用同一输入文件。