Installation¶

This page describes how to install DSF, configure a Python environment, and verify the installation.

Getting Started¶

Activate the Python environment used for your cosmology and forecasting work. For example, with conda:

conda activate <your-env-name>

Install the latest released version of DSF from PyPI:

python -m pip install lsst-desc-dsf

DSF requires Python 3.10 or newer.

Verify that Python can import DSF:

python -c "import dsf; print(dsf.__file__)"

The PyPI distribution is named lsst-desc-dsf, while the Python package is imported as dsf.

To work on DSF, clone the repository and navigate to its root directory:

git clone https://github.com/LSSTDESC/dsf.git
cd dsf

Install DSF in editable mode:

python -m pip install -e .

Editable installation allows changes made in src/dsf to take effect without reinstalling the package.

For development work, install DSF together with its development dependencies from the repository root:

python -m pip install --group dev -e .

Development should take place on a dedicated Git branch rather than directly on main.

Before starting substantial changes, open or find a GitHub issue describing the proposed work.

For example, create a documentation branch with:

git switch -c docs/installation-page

Prefer small, focused pull requests that are easy to review.

DSF uses tox to manage testing, linting, and documentation workflows.

List the available environments:

tox list

Run the test suite:

tox -m test

Run linting checks:

tox -e lint

Build the documentation:

tox -e docs

For detailed development guidelines, including code style, testing, documentation, and pull requests, see Contributing.