An Empirical Assessment of Polyglot-ism in GitHub

In this paper we study how the language cocktails are composed. How many languages are used in each software projects, which language types are used and which languages are typically used together. Our study was done on a sample of over 15,000 projects from the largest software forge, GitHub. The results show that many languages are used in each project: 96% projects employ at least 2 languages, over 50% employ at least two programming languages. Finally, there are strong relations between different languages: hence sets of languages tend to be adopted together