Rethinking Literate Programming in Statistics

Abstract

Literate programming is becoming increasingly trendy for data analysis because it allows the generation of dynamic-analysis reports for communicating data analysis and eliminates untraceable human errors in analysis reports. Traditionally, literate programming includes separate processes for compiling the code and preparing the documentation. While this workflow might be satisfactory for software documentation, it is not ideal for writing statistical analysis reports. Instead, these processes should run in parallel. In this article, I introduce the weaver package, which examines this idea by creating a new log system in HTML or LATEX that can be used simultaneously with the Stata log system. The new log system provides many features that the Stata log system lacks; for example, it can render mathematical notations, insert figures, create publication-ready dynamic tables, and style text, and it includes a built-in syntax highlighter. The weaver package also produces dynamic PDF documents by converting the HTML log to PDF or by typesetting the LATEX log and thus provides a real-time preview of the document without recompiling the code. I also discuss potential applications of the weaver package.

Seamless interactive language interfacing between R and Stata

“…Embedding languages in statistical software have followed the usual procedure of computer science without considering the practical demands of applied statistics. I argued that in the field of statistics, the analysis scripts are often developed interactively, which is usually neglected (Haghish 2016b,c), particularly in language interfacing for statistical software. Underscoring the importance of interactive workflow, I proposed that language interfacing in data analysis should allow interactive calls with seamless data communication.…”

On the importance of syntax coloring for teaching statistics

“…Several Stata packages that produce HTML or L A T E X files can benefit from the syntax highlighter for developing appealing educational materials. For example, the JavaScript and the L A T E X engines are already implemented in the (Haghish 2016c) and (Haghish 2016b) packages for developing dynamic analysis documents. Because of its simple installation and customization, can be easily added to websites, HTML documents, or L A T E X documents to facilitate beginners’ code comprehension and make educational materials more appealing for learners.…”

Software documentation with markdoc 5.0

“…After you install the repository, the command installs three dependency packages, which are specified in a file named . 3 The dependency packages are (Haghish 2016e), (Haghish 2016c), and (Haghish 2019a). The package includes the , , and commands, used for writing dynamic text, capturing and adding figures, and creating dynamic tables, respectively.…”