Installation steps

Installation steps#

1. Create and activate the examples environment#

We recommend using a dedicated Conda environment to ensure all native dependencies (HDF5, PROJ, GEOS, etc.) are available.

Warning

If you only want to run the notebooks, cloning the repository is sufficient.
If you plan to contribute to the code or documentation, we recommend forking the repository first and cloning your fork instead.

# Users
git clone https://github.com/OceanCruises/LAMTA_examples

# Contributors
#git clone https://github.com/<your-username>/LAMTA_examples

cd LAMTA_examples

conda env create -f environment.yml
conda activate lamta_examples

This installs the scientific stack required by the notebooks (numpy, scipy, xarray, netCDF4, cartopy, matplotlib, etc.).

2. Link your existing LAMTA installation to this environment#

If you are developing LAMTA and already have a local clone, install it in editable mode into the activated lamta_examples environment:

pip install -e /absolute/path/to/your/LAMTA

This reuses your existing working tree and ensures that any local changes to LAMTA are immediately visible to the notebooks.

If LAMTA is already installed in editable mode in another environment, you must still repeat this step for lamta_examples environment (editable installs are environment-specific).

3. Verify that LAMTA is correctly available#

python -c "import lamta; print('LAMTA import OK')"

If this fails, ensure that:

the correct environment is activated,
LAMTA was installed into this environment.

4. Open the notebooks#

You can use any Jupyter-compatible interface (JupyterLab, classic Jupyter, VS Code, etc.).

Make sure the lamta_examples environment is activated
Open a notebook from the notebooks/ directory
When prompted, select the lamta_examples Python kernel

For example:

VS Code: open a notebook → select kernel in the top-right corner
JupyterLab / Jupyter Notebook: choose the kernel when opening the notebook