Installation

Prerequisites

• Python ≥ 3.10
• PyADM1ODE (base package) — installed transitively. If not automatically
  available:

pip install git+https://github.com/dgaida/PyADM1ODE.git

• Optional: PyADM1ODE_calibration for the plant builder and historical
  data access.

Standard installation

git clone https://github.com/dgaida/PyADM1ODE_estimate.git
cd PyADM1ODE_estimate
pip install -r requirements.txt
pip install -e .

This installs the core dependencies: numpy, scipy, pandas, pythonnet, pyyaml, pyadm1.

Optional components

Deep learning (PINN, ensemble)

pip install -e ".[deep_learning]"

Adds torch and scikit-learn.

Development tools

pip install -e ".[dev]"

Adds pytest, pytest-cov, black, ruff.

Building the documentation locally

pip install -e ".[docs]"
mkdocs serve

The documentation is then served at http://localhost:8000.

Verification

python -m pytest tests/test_ukf.py tests/test_observation_model.py \
                 tests/test_state_vector.py tests/test_calibration_artifact.py

All tests should pass.

Troubleshooting

See Troubleshooting for common issues.