Types of Graphs - classifications

Models & Apps
#1

So much interesting:

https://docs.dgl.ai/tutorials/basics/4_batch.html

I was looking for smtg similar since years.

https://docs.dgl.ai/api/python/data.html#dgl.data.MiniGCDataset

Could you please other dataset to extend collection of types of graphs ad rationale of how graphs have been chosen ?

As a discussion, imagine a real world graph:

are these graphs considered “lego bricks” to describe a vast complexity of networks, or have been arbitrarily chosen?

I basically wonder if it would make sense to consider real world graphs as compositions of this preset.

Please advice or pros and cons.

My purpose is to classify structures.

#2

We just create the dataset out of common random graphs provided in networkx for a demo. In real world the graphs can be highly arbitrary, though you can probably consider common blocks like circles.