Graph Embedder

The Task

This task consists of creating an embedding of a graph. That is, a vector of features which can be used for a downstream task. The GraphEmbedder and GraphClassificationData classes internally rely on pytorch-geometric.

---

Example

Let’s look at generating embeddings of graphs from the KKI data set from TU Dortmund University.

We start by creating the TUDataset <https://pytorch-geometric.readthedocs.io/en/latest/_modules/torch_geometric/datasets/tu_dataset.html#TUDataset>. Next, we load a trained GraphEmbedder (from a previously trained GraphClassifier). Finally, we save the model. Here’s the full example:

import torch

To learn how to view the available backbones / heads for this task, see Backbones and Heads.