Skip to content

Dataset · rugds

Water distribution network graph time series

rugds/ditec-wdn

A massive spatial-temporal graph array with simulations of 36 water supply systems for Graph ML.

What's inside

Contains between 1B and 10B records describing 36 water distribution networks, with 1,000 unique scenarios generated for each. Each scenario is a time series of directed graphs with a fixed topology, where nodes (reservoirs, junctions, tanks) and edges store multi-dimensional time series.

What you can do with it

  • Training graph neural networks (GNNs) for infrastructure state forecasting
  • Modeling hydraulic processes and anomaly detection in infrastructure
Tasks
Graph ml, Time series forecasting
Languages
English
Size
1B–10B rows
Downloads this month
30.5K
Likes
4

How to load it

from datasets import load_dataset

ds = load_dataset("rugds/ditec-wdn")