Imported from losnleger/GreenAmpt-Nash-Muskingum (skills/greenampt-nash-muskingum/SKILL.md). Install upstream with npx skills add losnleger/GreenAmpt-Nash-Muskingum --skill greenampt-nash-muskingum. Copyright stays with the author.

GreenAmpt-Nash-Muskingum

Use this repository as a teaching and research prototype for lumped rainfall-runoff modeling. Do not present synthetic example-data results as professional validation.

Quick Workflow

From the repository root:

python -m pip install -e .
python generate_example_data.py
python calibrate.py dataset/example_data.csv -o params/calibrated_params.json --figures-dir outputs -a 150 --dt 1 -e 100 -p 80 --seed 42
python forecast.py dataset/example_forecast.csv -p params/calibrated_params.json -o outputs/forecast_result.csv --plot outputs/forecast_result.png

For a fast agent smoke check:

python skills/greenampt-nash-muskingum/scripts/smoke_check.py --epochs 2 --population 6

Use Real Data

Calibration CSV must include Rainfall_mmh and Observed_Flow_m3s; Time_h and Split are optional. Prefer explicit Split=train/test labels. Without labels, calibrate.py uses chronological --train-ratio.

Forecast CSV must include Forecast_Rainfall_mmh or Rainfall_mmh.

Read references/io-schema.md when creating or checking input files.

Validate Before Claims

Run at least:

python -m compileall .
python -m unittest discover -s tests
python skills/greenampt-nash-muskingum/scripts/smoke_check.py --epochs 2 --population 6

Report what data was used. If only the synthetic example data was used, say the result is a workflow smoke test, not real-basin validation.

Professional Boundaries

Read references/model-card.md before writing public descriptions, marketplace copy, papers, or repository summaries.

Required wording:

• Acceptable: "teaching/research prototype", "secondary development template", "synthetic example workflow".
• Avoid: "operational flood forecasting system", "engineering validated", "safety-critical decision system".
• Never use this repository alone for evacuation, dam operation, or regulatory decisions.

Release Hygiene

Before staging or publishing:

git status --short --ignored

Do not publish caches, outputs/, local params/, private Excel files, real project data, or credentials. Keep dataset/example_data.csv and dataset/example_forecast.csv only as synthetic demonstration data.