Contributing Overview

Event-LAB contributions should keep the shared pipeline simple: datasets define where data comes from, baselines define how a method runs, and config.yaml defines the shared experiment settings.

Before opening a pull request

1. Run the smallest command that exercises your change.

2. Check that outputs are written under output/.

3. Check that metrics are appended to output/eventlab_results.xlsx when a baseline is run.

4. Keep unrelated formatting and refactors out of the change.

Repository conventions

datasets/

Dataset YAML files and dataset loading helpers. A dataset YAML should be enough for Event-LAB to download, format, and generate frames for a sequence.

baselines/

Baseline wrappers and baseline-specific YAML files. A wrapper should adapt one method to Event-LAB’s reference/query evaluation flow.

docs/

Read the Docs source files. Keep pages short, command-focused, and close to the current code behavior.

config.yaml

The shared experiment configuration used by normal runs and batch runs.

Minimal test command

Use a small known run when checking documentation or pipeline changes:

pixi run eventlab sparse_event brisbane_event sunset2 sunrise

For documentation-only changes, also build the docs:

pixi run sphinx-build -b html docs docs/_build/html