Development

Local Setup

python3 -m pip install -e ".[test,export]"

Run tests:

.venv/bin/python -m pytest -q

Repository Layout

src/pyoncoplot/                         package code
tests/                                   pytest suite
docs/                                    Markdown documentation
plans/                                   implementation plans
python_refactor_goal_sources/            gallery scripts and training source tree
python_refactor_goal_sources/config.yaml config-driven gallery run definitions
python_refactor_goal_sources/make_oncoplots.py direct oncoplot() gallery renderer
python_refactor_goal_sources/goal_plots/ numbered source/reference plots
python_refactor_goal_sources/fuc_sources/ fuc source scripts and fixture rebuild helper
python_refactor_goal_sources/syntheitic_goal_data/ deterministic TSV/JSON gallery inputs

Regenerate Gallery Inputs

python3 python_refactor_goal_sources/generate_synthetic_inputs.py

This rewrites deterministic non-fuc TSV/JSON files under:

python_refactor_goal_sources/syntheitic_goal_data/

The generator reads output filenames and deterministic non-fuc generation settings from gallery_params.input_files and gallery_params.synthetic_inputs in python_refactor_goal_sources/config.yaml.

The AML fuc fixtures are rebuilt separately after downloading the upstream fuc-data source files listed in python_refactor_goal_sources/fuc_sources/manifest.json:

python3 python_refactor_goal_sources/fuc_sources/rebuild_fuc_fixtures.py --source-dir /path/to/fuc-data

Regenerate Gallery Images

python3 python_refactor_goal_sources/recreate_gallery.py
python3 python_refactor_goal_sources/recreate_gallery.py --style comparison --preset brca_large

Generated gallery images are tracked when intentionally updated.

Adding A Gallery Preset

1. Add or extend synthetic input settings in python_refactor_goal_sources/config.yaml, or fuc-derived inputs in python_refactor_goal_sources/fuc_sources/rebuild_fuc_fixtures.py.

2. Commit the generated TSV/JSON inputs.

3. Use renderer: oncoplot; active gallery goals must be expressible through public oncoplot() params.

4. Add the named run to python_refactor_goal_sources/config.yaml under gallery_params.plot_runs.

5. Add or update tests in tests/test_gallery.py.

6. Document the preset in Gallery.

The generic oncoplot renderer reads params.oncoplot from config and passes it to the public API. Keep runtime choices such as genes, metadata tracks, legend controls, expected dimensions, output names, and save DPI in YAML.

Documentation Changes

Docs are Markdown-first. Keep examples runnable from the repository root unless the page says otherwise.

After editing docs:

.venv/bin/python -m pytest -q

The docs test checks that required pages exist and that local Markdown links resolve.