You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It would be nice to actually show a crude drawing of the graph around the current node. It might also be nice to see a crude drawing of a clustering of the whole graph, since that would tell us some useful things about the overall structure.
We can get a neighborhood around the current node straightforwardly in networkx:
We could use graphviz to get a nice layout of the neighborhood, and then discretize that layout to the character grid. Graphviz seems to have a python wrapper which is in turn integrated with networkx:
Graphviz (I think?) wants to know, for pairs of nodes, how far apart those nodes want to be. So, we would need to come up with some heuristic for that. It looks like networkx will have something reasonable in its algorithms:
It would be nice to actually show a crude drawing of the graph around the current node. It might also be nice to see a crude drawing of a clustering of the whole graph, since that would tell us some useful things about the overall structure.
We can get a neighborhood around the current node straightforwardly in networkx:
We could use graphviz to get a nice layout of the neighborhood, and then discretize that layout to the character grid. Graphviz seems to have a python wrapper which is in turn integrated with networkx:
Graphviz (I think?) wants to know, for pairs of nodes, how far apart those nodes want to be. So, we would need to come up with some heuristic for that. It looks like networkx will have something reasonable in its algorithms:
I'm not sure which algorithm would be best, but we should probably just play with different ones until we find one that feels right.
The text was updated successfully, but these errors were encountered: