This repository contains the Water-Futures Team implementation of the Battle of the Water Demand Forecasting (BWDF) competition.
The Battle of the Water Demand Forecasting (BWDF) was a competition organized in the context of the 3rd Water Distribution Systems Analysis and Computing and Control in the Water Industry (WDSA-CCWI) joint conference held in Ferrara, Italy in 2024 (Alvisi et al., 2025).
This repository is our open-source implementation created to fulfill the Water-Futures mission of making water research tools freely available.
This repository provides an easy-to-use Python package that allows researchers to:
- Quickly load the competition data
- Evaluate models using the original competition workflow
You can install the package using pip:
pip install wf4bwdfAll required dependencies are installed automatically with pip. See pyproject.toml for details.
The project has a minimal dependency, from Python 3.9 and pandas above 2.1 it should work.
A lightweight implementation of this package is also available through the WaterBenchmarkHub (Artelt et al., 2025), which includes only the core functions to load data in memory and evaluate forecasts.
The function load_complete_dataset provides access to DMA inflows and weather data from the supplementary information of Alvisi et al., (2025) and also calendar information and other problem metadata readily available in machine-readable format.
import wf4bwdf as bwdf
dataset = bwdf.load_complete_dataset()
# Print DMA description
print(dataset['dma-properties']['Description'])
# Plot DMA 3(C) inflow
dma_c_inflow = dataset['dma-inflows']['DMA 3']
# plot the seriesEvaluate the forecast following the competition requirements means that the evaluate function works only if the forecast is a complete prediction of one the original evaluation weeks and of at least 1 DMA.
import wf4bwdf as bwdf
import pandas as pd
for iteration in range(1,5):
# Load the data for that iteration (no leak of future information) using letters instead of the numbers (e.g., 'DMA C')
dataset = bwdf.load_iteration_dataset(iteration, use_letters_for_names=True)
# Compute your forecast: previous week
forecast = dataset['dma-inflows'].iloc[-168:]
forecast.index = forecast.index + pd.Timedelta(weeks=1)
# Evaluate the forecast
results = bwdf.evaluate(forecast)
# Should have returned a series with the combination 'Evaluation week', DMA, and BWDF performance indicators')
print(results)See CONTRIBUTING.md.
This project is licensed under the MIT License. See the LICENSE file for details.
The data used in this project have been downloaded from the open access work of Alvisi et al. (2025). Those data are available under the terms of Creative Commons Attribution 4.0 International license, https://creativecommons.org/licenses/by/4.0/.
If you use this implementation of the BWDF in your research, please cite the original paper that you can find here: Battle of the Water Demand Forecasting paper.
Use the "Cite this repository" button in the About section (right sidebar) to copy the citation in your preferred format.
@article{Alvisi_Battle_of_Water_2025,
author = {Alvisi, S. and Franchini, M. and Marsili, V. and Mazzoni, F. and Salomons, E. and Housh, M. and Abokifa, A. and Arsova, K. and Ayyash, F. and Bae, H. and Barreira, R. and Basto, L. and Bayer, S. and Berglund, E. Z. and Biondi, D. and Boloukasli Ahmadgourabi, F. and Brentan, B. and Caetano, J. and Campos, F. and Cao, H. and Cardona, S. and Carreño Alvarado, E. P. and Carriço, N. and Chatzistefanou, G.-A. and Coy, Y. and Creaco, E. and Cuomo, S. and de Klerk, A. and Di Nardo, A. and DiCarlo, M. and Dittmer, U. and Dziedzic, R. and Ebrahim Bakhshipour, A. and Eliades, D. and Farmani, R. and Ferreira, B. and Gabriele, A. and Gamboa-Medina, M. M. and Gao, F. and Gao, J. and Gargano, R. and Geranmehr, M. and Giudicianni, C. and Glynis, K. and Gómez, S. and González, L. and Groß, M. and Guo, H. and Habibi, M. N. and Haghighi, A. and Hammer, B. and Hans, L. and Hayslep, M. and He, Y. and Hermes, L. and Herrera, M. and Hinder, F. and Hou, B. and Iglesias-Rey, A. and Iglesias-Rey, P. L. and Jang, I.-S. and Izquierdo, J. and Jahangir, M. S. and Jara-Arriagada, C. and Jenks, B. and Johnen, G. and Kalami Heris, M. and Kalumba, M. and Kang, M.-S. and Khashei Varnamkhasti, M. and Kim, K.-J. and Kley-Holsteg, J. and Ko, T. and Koochali, A. and Kossieris, P. and Koundouri, P. and Kühnert, C. and Kulaczkowski, A. and Lee, J. and Li, K. and Li, Y. and Liu, H. and Liu, Y. and López-Hojas, C. A. and Maier, A. and Makropoulos, C. and Martínez-Solano, F. J. and Marzouny, N. H. and Menapace, A. and Michalopoulos, C. and Moraitis, G. and Mousa, H. and Namdari, H. and Nikolopoulos, D. and Oberascher, M. and Ostfeld, A. and Pagano, M. and Pasha, F. and Perafán, J. and Perelman, G. and Pesantez, J. and Polycarpou, M. and Quarta, M. G. and Que, Q. and Quilty, J. and Quintiliani, C. and Ramachandran, A. and Reynoso Meza, G. and Rodriguez, V. and Romano, Y. and Saldarriaga, J. and Salem, A. K. and Samartzis, P. and Santonastaso, G. F. and Savic, D. and Schiano Di Cola, V. and Schol, D. and Seyoum, A. G. and Shen, R. and Simukonda, K. and Sinske, A. and Sitzenfrei, R. and Sonnenschein, B. and Stoianov, I. and Tabares, A. and Todini, E. and Tsiami, L. and Tsoukalas, I. and Ulusoy, A.-J. and Vamvakeridou-Lyroudia, L. and van Heerden, A. and Vaquet, J. and Vaquet, V. and Wallner, S. and Walraad, M. and Wang, D. and Wu, S. and Wu, W. and Wunsch, A. and Yao, Y. and Yu, J. and Zanfei, A. and Zanutto, D. and Zhang, H. and Ziebarth, M. and Ziel, F. and Zou, J.},
doi = {10.1061/JWRMD5.WRENG-6887},
journal = {Journal of Water Resources Planning and Management},
number = {10},
title = {{Battle of Water Demand Forecasting}},
volume = {151},
year = {2025}
}
This repository is mantained on initiative of the Water-Futures team and approved by the competition organizers.
Explore more of our projects on GitHub and learn about our team and work on our website.
Water-Futures has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 951424).