Combination Code

This is a generic averaging code to combine an arbitrary set of (experimental) results under consideration of the statistical errors and their correlations, and systematic errors and correlations, both between the parameters of one measurement and different measurements. The code documentation providing detailed instructions to the example code is available here.

A tutorial is available here

Installation

Install with `pip in a virtual environment

python3 -m venv coco-env
cd coco-env
source bin/activate
python -m pip install combination-code

Doucumentation and examples are available in share/combination-code

The Python package is exposed as:

from combination_code import ResultList, ResultSet, readResultListJSON

Development

Install the package in editable mode with development tooling:

python3 -m venv coco-env
cd coco-env
source bin/activate
git clone ssh://git@gitlab.cern.ch:7999/hflav/shared/combination-code.git
cd combination-code
pip install -e .[dev]

Run all tests (unit + system):

pytest

Test Suite

The repository includes two test layers:

• Unit tests in tests/unit/ for core APIs in ResultSet, ResultList, and IntercorrelationMaps.
• System tests in tests/system/ that execute tutorial examples from tutorial/Example1, tutorial/Example2, and tutorial/Example3 and validate expected output artifacts.

System tests force a non-interactive Matplotlib backend (Agg) so they run reliably in CI.

Continuous Integration

GitHub Actions workflow build.yml now:

• Runs the test suite on Python 3.10 and 3.14.
• Produces coverage output via pytest-cov.
• Uploads coverage.xml as a CI artifact.

Build and release (PyPI)

This is done via Gitlab CI. Simply create a new release with the for vX.Y.Z format following semantic versioning.