Przemysław Nowak
supervisor: Grzegorz Siudem
In the studies of scale-free networks, where degree distribution follows a power law, there are dosen of methods to analyse them. Most of them focus on heavy-tailed distribution, where the main goal is to compute power law exponent.
In our case, we switch perspective into rank-size distribution, which is a distribution of size by rank, in decreasing order. From mathematical point of view, it's nothing but inverse survival function of given random variable. However, this allow us to recreate new methods to fit parameters and compute goodnes of fit.
This way we perform mentioned methods for the networks of commits from GitHub. Recently we discover that DGBD (Discrete Generalized Beta Distribution) is a good model to predict numer of commits in various repositories from GitHub – from small to big, from open source to closed source, and many more.