Closed cbouss closed 1 year ago
Easy part is implemented.
Could use https://scikit-learn.org/stable/modules/clustering.html#affinity-propagation or https://scikit-learn.org/stable/modules/clustering.html#spectral-clustering This SO post has an example from a networkx graph: https://stackoverflow.com/questions/23684746/spectral-clustering-using-scikit-learn-on-graph-generated-through-networkx
We could identify group based on github/gitlab/... organization. This way we could support query build order for a group (example: build order of jupyter related packages).
More complex, but we could also identify groups based on the dependency graph, by identifying clusters. Gephi would do that, but there are likely some better suited python tools readily available