Introduce new plots that visualise a GNN and its input graphs
Graph Neural Networks (GNNs) have become increasingly popular in recent years, and it is important to have tools to visualise them and the input graphs they operate on. Currently, we only have limited visualisation tools for GNNs, and it would be beneficial to have new plots to help users better understand the structure and behaviour of GNNs.
To visualise a GNN, we can create a plot that shows the structure of the GNN, including the layers, connections between layers, and the input and output of each layer. We can create plots showing each graph's nodes, edges, and features to visualise the input graphs.