One example (save in test.csv and load undirected):
The network diameter is (correctly) calculated as 2. But the heatmap and shortest path features paint/calculate longer distances than 2 on some nodes/pairs. Just use heatmap on multiple nodes and some of them will show a max distance of 3.
It looks like the implementation of Dijkstra's algorithm is incorrectly returning a distance of 3 for certain nodes. Possibly related to the closed issue #472?
Fix for issue #630
Hi oxling, thanks for the report and fix :)