The interactive visualizations on this page are based on the excellent Viz project. Donne Martin, the creator of Viz, mined GitHub data, extracting the metadata for all repositories created in 2015 with >100 stars. He has shared a number of different visualizations based on this data.

Donne has made the underlying data available, which has allowed me to create a similar set of visualizations using D3 and d3fc. As one of the creators of d3fc, I am using this as an oppurtunity to demonstrate the power of the component-based approach, which allows you to create bespoke charts and visualizations from a range of building blocks.

The above charts plots all of the ~7,000 repos in the dataset as a scatter chart of stars vs. forks. The majority of the repositories sit beneath the stars = forks line, indicating that the majority of repositories have more stars than forks. If you mouse-over the legend you can see that the distribution for each language.

There are some interesting patterns, for example JavaScript projects receive a lot of stars compared to forks, whereas C#, Java and C++ received a much higher ratio.

Of course this relationship is much easier to see when the forks-per-star ratio is plotted directly:

The above makes use of d3fc's newly added boxplot series component.

The next chart looks at whether projects belong to individuals (users), or organizations. You can see that there is a tendency for projects to be managed by organizations, however there is quite a bit of variability based on language.