epymodelingsuite

Models and calibrators for forecasting and scenario modeling. A codebase for large workloads with epydemix and expanded functionality.

Installation

To install directly from GitHub (recommended for normal usage), run pip install git+ssh://git@github.com/mobs-lab/epymodelingsuite.git. You must have an SSH keypair for your machine enrolled with GitHub and have access to the repo.

Development

Environment Setup

Option 1: uv

make install-dev    # Creates isolated .venv in project directory
uv run pytest       # Run commands via uv

uv manages its own .venv. There is no need to activate anything.

Option 2: pip/conda

conda activate myenv    # Activate your environment first
make install-dev-pip    # Installs locked dependencies into active env
pytest                  # Run commands directly

Installs into whatever environment is currently active (conda, venv, system Python).

Lock Files

• uv.lock is the single source of truth; requirements.lock.txt is exported from it.
• Run make lock to update after changing pyproject.toml.

Jupyter Kernels

If you installed with uv (option 1) and want to use Jupyter notebooks:

• Using notebooks in VSCode: VSCode auto-detects the .venv directory. Select it as the Python interpreter/kernel.
• Running JupyterLab from uv: Start new JupyterLab with uv run --with jupyterlab jupyter lab. The kernel is available automatically.
• Using an existing JupyterLab installation: If you have JupyterLab set up separately, register the kernel manually:

  uv run python -m ipykernel install --user --name epymodelingsuite --display-name "epymodelingsuite"

Documentation