PyTorch Geometric Dataset

As explained in Graph Neural Networks in Action

A PyTorch Geometric base class for defining datasets and their raw-data processing, filtering, transformation, and retrieval behavior.

Use a PyTorch Geometric Dataset to turn raw graph files into repeatable data objects that support preprocessing, splitting, batching, and model ingestion. The Dataset and InMemoryDataset classes organize this pipeline, while Data represents each graph. section 8.3.4

Custom social-network dataset

A process() implementation reads an edge list with NetworkX, maps named nodes to integer indices, constructs edge_index, x, y, and masks, wraps them in Data, and collates the result.

Figure 8.15 Steps to preprocess data in PyG. From raw files, there are essentially two paths to prep data for ingestion by a PyG algorithm. The first path, shown here, directly creates an iterator of data instances, which is used by the dataloader. The second path mimics the first but performs this process within the dataloader class.

Use the preprocessing figure to compare direct iteration from raw files with preprocessing performed within a dataloader. section 8.3.4

How Graph Neural Networks in Action covers it

How deep it goes

Explains initialization, file declarations, downloading, graph conversion, masking, collation, and processed-data loading without detailing Dataset internals.

Examples

Demonstrates a NetworkX edge-list conversion into indexed tensors, Data attributes, masks, a custom dataset, and a batched loader.

What it assumes

Assumes Python, PyTorch tensors, NetworkX graph data, and basic PyG Data objects; demonstrates node indexing and masks directly.

What you can do after

Build it from scratch by implementing a custom InMemoryDataset that processes raw graph files into collated Data objects.

How it connects

Connects Dataset with Data, InMemoryDataset, DataLoader, preprocessing, graph indexing, train/test masks, and batching.

liveBookmanning.com

Read Graph Neural Networks in Action on liveBook Get the book at manning.com

Related concepts

• PyG InMemoryDatasetincludes · A PyTorch Geometric dataset class that stores its processed data in memory.
• datasetkind of · A dataset is an organized collection of related data items, records, or observations treated as a unit for storage, processing, analysis, or computation.

This explainer was generated by Manning's knowledge graph from the text of Graph Neural Networks in Action. The section links point to the parts of the book it draws on.