This project is an analysis using a case study gitlab project: Tiki Project
- It produces results from Data (log) extraction and collection, Data clean-up, Data Categorization, Data Visualisation of developer contributions to the project.
- I use Leiden Algorithm and temporal network parameters by defining edge data with source, target, and unique identifiers to view the contibution in different time snapshots.
- Overall, this result show a supervised machine learning model and pattern useful for network groupings and developer recommendation for bug fixes
The data is then extracted from the local Git repository using the GitExtractor.
- Once extracted, a GitExtractor method can be called to return data in a Pandas DataFrame.
Overall, the network comprises of 174 Nodes and 28021 Edges.
- The thickness of the edge represent the contribution weight and the red nodes represent the bug related instances: [FIX], [ENH], [MOD], [REF], [NEW], [REL], [UPD], [KIL], [ADD], [SEC], [CSS], [UI], [SVN].
- The sequence of contributions points out some key developers that have shown sustained contribution over time.
- Temporal networks are network representations that flow through time. They give a view of the network as it develops over time, taking snapshots at a few key moments over the course of its timespan
- Degree, betweenness and centrality.
Ranking: Using Node Degree (15 <= x <= 23)
Average degree:0.065 Graph Density:0 Leiden algorithm: 0.984 Avg. Weighted Degree: 0.549