AVYield - Analysis and Visualization of Crop Yield Trials

AVYield is a digital interactive decision dashboard for crop yield trials, enabling near-real-time exploration and comparison of multi-year, multi-location crop trial data. It overlays farmers' own data with formal trial results to guide genotype and management decisions.

Try it out at AVYield!

---

Paper & Citation

• Link: A digital interactive decision dashboard for crop yield trials
• Citation:

@article{CISDELI2025110037,
  title = {A digital interactive decision dashboard for crop yield trials},
  journal = {Computers and Electronics in Agriculture},
  volume = {231},
  pages = {110037},
  year = {2025},
  issn = {0168-1699},
  doi = {https://doi.org/10.1016/j.compag.2025.110037},
  url = {https://www.sciencedirect.com/science/article/pii/S0168169925001437},
  author = {Pedro Cisdeli and Gustavo {Nocera Santiago} and Carlos Hernandez and Ana Carcedo and P.V. Vara Prasad and Michael Stamm and Jane Lingenfelser and Ignacio Ciampitti},
  keywords = {Digital tool, Yield data, Crop genotypes, Visualization, Data repository},
}

---

Features

• Filter and visualize trial data by crop, location, year, genotype, and water regime.
• Overlay custom farmer data on trial benchmarks.
• Perform on-the-fly genotype comparison analyses.
• Download datasets and export plots.

---

Installation & Local Setup

Prerequisites

• Python 3.8 or higher
  • Tested in Python 3.11
• pip

Clone Repository

git clone https://github.com/cisdeli/AVYield.git
cd AVYield

Install Dependencies

pip install -r requirements.txt

Run Locally

Edit app.py to update the server startup call:

app.run_server(host='0.0.0.0', port=8050)  # local-host
# Instead of:
# app.run_server(debug=True, host="0.0.0.0", port=8080,
#                use_reloader=False)  # Docker

Then run:

python app.py

Open your browser to http://localhost:8050 to access the dashboard.

---

License

This project is licensed under the CC BY 4.0 License.

---

Acknowledgements

Developed by Pedro Cisdeli and collaborators. For full details, see the paper linked above.