Installation#

Requirements#

  • Python >= 3.10

Runtime dependencies are declared in pyproject.toml with minimum compatible versions and are resolved automatically when installing via pip.

User installation#

Install the latest release from PyPI:

python -m pip install gedidb

Optional features#

Install optional components using pip “extras”:

  • Full stack (IO + viz + performance + logging)

    python -m pip install "gedidb[full]"

  • Mix-and-match individual extras (advanced):

    • accel — performance helpers (bottleneck, numba, etc.)

    • io — cloud/filesystem and raster IO (s3fs, zarr, rasterio, pyproj, etc.)

    • viz — plotting (matplotlib, cartopy, seaborn, etc.)

    • debug — rich logging

    Example:

    python -m pip install "gedidb[io,viz]"

External system libraries#

Some geospatial backends rely on native libraries:

  • GDAL / PROJ / GEOS — Recommended to install via a conda-forge environment if you are not already set up with these on your system:

    mamba create -n gedidb -c conda-forge python>=3.10 gdal proj geos
mamba activate gedidb
python -m pip install "gedidb[full]"

  • wget — used by EarthDataAuthenticator for downloads. This is preinstalled on most Linux/macOS; on Windows you can install via winget or choco.

Development setup#

Clone and install an editable development environment with tooling:

git clone https://github.com/simonbesnard1/gedidb.git
cd gedidb

# (optional) create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate

python -m pip install -U pip wheel
# editable install with dev tools (pytest, ruff, mypy, pre-commit, etc.)
python -m pip install -e ".[dev,full]"
pre-commit install

Install latest development snapshot directly from GitHub (no clone):

python -m pip install "git+https://github.com/simonbesnard1/gedidb.git"

Testing#

From the project root:

pytest

With coverage:

pytest --cov=gedidb --cov-report=term-missing

If you maintain integration tests that require credentials or large datasets, mark them with @pytest.mark.integration and run selectively:

pytest -m "not integration"