Network Statistics

[simisimon]

I want to investigate the evolution of the configuration space in modern software development.
To this end, we need to calculate statistics during the the analysis of the commit history of a software system.

Statistics of interests are:

• size of the network in terms of the number of nodes
• total number of each node type
• number of links
• how often are values changed
• which configuration file are changed
• how often are links added/destroyed
• what is the value range of options

We probably cannot calculate all statistics in the network. In this case, we need to prepare and store the necessary data somewhere (e.g., csv file), so that we can calculate those statistics later, for example, with an external script.

[simisimon]

Since we want to analyze the commit history of the software system, we have to calculate these numbers for each commit. However, since configuration changes are not made in every commit, it would make sense to analyze only those commits in which configuration files are changed. This way, we would decrease the number of csv files we have to create.