Monash Time Series Forecasting Repository

Provider
Rakshitha Godahewa, Christoph Bergmeir, Geoffrey Webb, Rob Hyndman, Pablo Montero-Manso (Monash University)
License
CC-BY-4.0 (source)
Size
36 files
Stored at
Hugging Face (https://huggingface.co/datasets/Monash-University/monash_tsf)
Access
Free and public. No account with the data provider is needed.
Category
Machine Learning & AI
Source
https://arxiv.org/abs/2105.06643

About Monash Time Series Forecasting Repository

The Monash Time Series Forecasting Repository, a benchmark archive of 30+ real-world time series datasets from many domains, introduced at the NeurIPS 2021 Datasets and Benchmarks track to give global forecasting models a common evaluation ground. Collections include Australian electricity demand, London smart meters, solar power at 10-minute and 4-second resolution, wind and weather, traffic and pedestrian counts, tourism (monthly, quarterly, yearly), M1, M3 and M4 competitions, NN5 cash withdrawals, CIF 2016, FRED-MD macroeconomic series, hospital and car parts demand, Kaggle web traffic, KDD Cup 2018 air quality, COVID deaths, rideshare, sunspots, Saugeen river flow, temperature and rain, and Bitcoin, several with variants with and without missing values. Listing data/ (measured 2026-11-14) shows 32 zip archives in the repository's .tsf format; the repository uses about 24.5 GB. Load configurations with Hugging Face datasets through the bundled monash_tsf.py loading script or read the .tsf files with the converter from the Monash GitHub repository and pandas or GluonTS. Licensed CC BY 4.0; cite Godahewa et al., Monash Time Series Forecasting Archive (2021).

How to load Monash Time Series Forecasting Repository in Python

from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="Monash-University/monash_tsf", filename="data/australian_electricity_demand_dataset.zip", repo_type="dataset")
print(path)

Data structure

Files

Example file: data/australian_electricity_demand_dataset.zip

Analyze this dataset without downloading it

On Scigantic, Monash Time Series Forecasting Repository opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Machine Learning & AI datasets