Language popularity on GitHub

An exploration of language popularity on Github.

My submission to a university course in data analysis and visualisation. I wanted to explore if there was any correlation between the features of any given programming language and the amount of bugs in the software built using that language.

While the analysis gave no conclusive evidence, and similar studies also find no such correlation, I learned a lot from this experience. It was the first time I had ever done any data analysis, and use R, and GraphQL. I decided to gather my own dataset as I noticed that publicly available datasets from GitHub were quite outdated. I wrote a few utilities in R using httpr to call the GitHub GraphQL API however many times was required to gather the amount of data I wanted. My final dataset contained data from 100 000 repositories. I used this data to analyse popularity of languages, language features, and to analyse if there was any correlation between language or language features and the amount of bug issues on a repository.

The analysis is quite naive in many areas and doesn't account for many factors, but I still find it very interesting to analyse and interact with. The most interesting part for me is definitely the analysis of popularity of language features and the trends in language popularity. Most charts are interactive which allows you to explore the results yourself.

The report was generated using RMarkdown compiled to HTML with some inline CSS and HTML. The interactive charts are made using Plotly.js.

You can check out the report at: VIS3000.fuglesteg.net

The source code and datasets are available on my GitHub