Updated the Megatrail so that it uses 1 of 2 separate methods of determining community size and number, namely, either directly defining the number of communities to use, or changing the number of communities by affecting resolution. The first version of this used resolution, which I notice is a tad blockier and more localized. By contrast, the direct method seems to produce shapes which are, to my eye, more organic.
The other thing to note is that using different resolutions produces different organizations entirely (although some communities are remarkably resilient to change; e.g., the community running from Elgin to Woodstock seems to be pretty universal across all maps), whereas the community number method uses the same overall organization, but cuts the communities into smaller chunks.
The other thing to note is that using different resolutions produces different organizations entirely (although some communities are remarkably resilient to change; e.g., the community running from Elgin to Woodstock seems to be pretty universal across all maps), whereas the community number method uses the same overall organization, but cuts the communities into smaller chunks.
For those who are interested, this is the method I'm using: https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.community.modularity_max.greedy_modularity_communities.html I'll be reading up more on this material today to see if I can better explain the change in shapes between methods.