Eugenio Angriman
Eugenio Angriman
This is a continuation of #803
I recently used `parallelForEdges` to generate random weights on large graphs (1-2 billion edges). On a 36-cores machine, the performance was very poor (it took >1h) due to bad load...
If we pass a directed weighted graph `Gd` with bidirectional edges such as `(u, v, w1)` and `(v, u, w2)` with `w1 != w2` to `GraphTools::toUndirected()`, the resulting undirected graph...
Some NetworKit classes inherit from multiple classes like `Algorithm`, `DynAlgorithm` or `Centrality`. For example, [DynTopHarmonicCloseness](https://github.com/kit-parco/networkit/blob/5cc74b434b4891ea53a278777a3699a26bedf379/networkit/_NetworKit.pyx#L6550l) extends both `Algorithm` and `DynAlgorithm`. However, when it comes to Cython there is no easy-way...
As recently reported in the mailing list, the following instruction outputs an undirected graph: ``` G = nk.graphio.readGraph('', nk.Format.EdgeListSpaceOne, directed=True) ``` While the output of: ``` G = nk.graphio.readGraph('', nk.Format.EdgeList,...
https://github.com/kit-parco/networkit/blob/00d28909d3e8833d35c0aa9c0d97cb9b3b7a0940/networkit/cpp/generators/ClusteredRandomGraphGenerator.cpp#L29 This assertion is not safe since it is not guaranteed that every subset will contain a node (especially if `k` approaches the number of nodes).
This completes #803, all cpp files in NetworKit are now clang-formatted
- Fix a potential bug in PLP (a variable was updated non-atomically in a parallel loop) - Replace `std::map` with `std::unordered_map` because the order of the elements is not needed...
Hi, I've noticed that the NetworKit shortest-path benchmark is executed with ``` distance.BFS(g, node_index).run() ``` However, with this API NetworKit will also store _all_ shortest paths from `node_index` to all...