Energy, Economics & Replication
McMaster University, Department of Economics Working Paper No. 2017-02
15 Pages Posted: 15 Feb 2017
Date Written: February 13, 2017
This article outlines recent developments in Markdown scripting languages that facilitate the production of replicable, publication quality, research. The approach is similar to that achieved by using, say, Sweave, R and LaTeX, but is written instead in simple Markdown syntax and not tied to any particular output format (e.g., MS Word) nor computational language (e.g., Python). The computational component can be written in C++, Python, SQL, Stan, Bash, or R by way of example. The Markdown script is seamlessly converted to any one of a number of output formats. The output format is essentially an afterthought, and could be rendered as a PDF (LaTeX or Beamer presentation), MS Word, HTML, EPUB, or gitbook document, by way of illustration. Conversion of the Markdown script to the desired output format is performed by pandoc (a universal document converter). These tools can dramatically reduce the amount of time required to complete a research project that can be trivially replicated. Recent enhancements to RStudio streamline the entire process of output format generation via a simple click of an icon or keystroke shortcut (the minimum requirement is R). Replicability is guaranteed by using the checkpoint package in R. This article was written using Markdown.
JEL Classification: C88
Suggested Citation: Suggested Citation