Detecting Outliers in Premise Operations (dopo)

About dopo

The Premise Validation Project introduces the Python package dopo (Detecting Outliers in Premise Operations), a toolkit for evaluating the consistency of background life cycle inventories and identifying anomalies that could affect LCA results.

It is essentially a wrapper around the Brightway2 bw2analyzer library, combined with a Dash web application for interactive exploration of datasets and results.

---

🧪 Features

• Detect structural and numerical outliers in ecoinvent-based inventories.
• Interface with Brightway2 and Premise-modified databases.
• Filter or classify activities based on CPC, ISIC, or custom sector definitions.
• Assess differences in environmental impacts from method-to-method or database-to-database.

---

📊 Dash Web Application

dopo includes a Dash-based interactive app for exploring datasets and visualizing environmental impact scores across projects.

App Features

• ✅ Load and switch between multiple Brightway projects
• ✅ Select one or more background databases
• ✅ Choose a dataset grouping: Sectors, CPC, or ISIC
• ✅ Filter datasets with a search bar
• ✅ (Optional) Exclude market activities using a checkbox
• ✅ Select one or more impact assessment methods
• ✅ View either total impact scores or contribution plots
• ✅ Interactive dropdowns to explore results by sector and method
• ✅ Run calculations and visualize scores in seconds

Installation

Important

You need to install dopo in a Python environment with brightway (2 or 2.5).

You can install dopo via from Anaconda:

$ conda install romainsacchi::dopo

Or via [pip] from [PyPI]:

$ pip install dopo

Important

For Mac users with an ARM chip, you need to have scikit-umfpack as well as numpy<=1.24.4, otherwise it results in slow calculations. This is not specific to dopo but to brightway in general.

How to run the app

You can launch the app directly from the terminal using the following command:

dopo-dash

Documentation

https://dopo.readthedocs.io/en/latest/

Contributing

Contributions are very welcome. To learn more, see the Contributor Guide.

License

Distributed under the terms of the MIT license, dopo is free and open source software.

Issues

If you encounter any problems, please file an issue along with a detailed description.

Maintainers

• Romain Sacchi, PSI
• Caroline Friedberger, NTNU

Support

Feel free to contact Romain Sacchi if you have any questions or need help.