These are some notes from while I was going through the post Exploring and Analyzing Network Data with Python.

When considering triadic closures and transitivity: is there a way of getting a list of all possible triangles? This is one thing that I'd love to use the graphs/networks for: surfacing connections that I could make but have yet to consider.

Is there ways of considering transitivity across clusters, such as regarding whole clusters as simply nodes in a fork? (Admittedly I'm asking this before getting to any discussion of cluster identification.) It'd be cool if one could develop a graph of all skateboard tricks and their various sequences of execution, then highlight ways that the moves are and are not being used (or in which order or direction, as with directed graphs), to surface new trick ideas.

You can use the built-in function sorted() to sort a dictionary by its keys or values and find the top twenty nodes ranked by degree.


Very dense networks are often more difficult to split into sensible partitions [clusters]