It may make more sense to do this on a figure by figure basis and have notebooks to demonstrate this functionality. Some figures we'd need to be able to reproduce are:
Heatmaps - these are those U-plots. need this for both the uniform and triangle weighted diffusions
NCPs - epidemic and pagerank versions
missed sets figure - note this requires both diffusion data and ppr ncp data
Figure showing lambda1 versus average degree vs lambda1 - can probably just cache that and send data as-is
The plotting files are especially convoluted and need to be cleaned up. We need the following: