Building and installing

Building and installing#

This page explains how to install, build from source, and test the project. For most users, a conda-based environment (Miniforge or Mambaforge) is recommended because it reliably provides all native dependencies.

Note

Windows is supported, but only via a conda-based environment. Building against system-wide dependencies on Windows is not supported.

Building with Miniforge3#

Important

If you are on Windows, use the conda-based workflow below. Other installation methods on Windows are not supported.

To install the prerequisites and set up the build environment using conda, follow these steps:

  1. Install Miniforge3 from the official source: conda-forge/miniforge

  2. Clone the repository:
    git clone https://github.com/CNES/aviso-lagrangian.git
  3. Navigate to the project directory:
    cd aviso-lagrangian
  4. Create the build environment using the provided environment file:
    mamba env create -f conda/environment.yml

    This will install all necessary dependencies for building, testing, and generating documentation.

  5. Activate the environment:
    mamba activate lagrangian
  6. To install this software in the created environment, run:
    pip install .

Building from Source#

Warning

On Windows, building from source outside a conda environment is not supported. Use the conda-based installation described above.

This section explains how to compile the code, install it, and run the various scripts.

This software uses Python’s Setuptools for building and installation. For more details, refer to the official documentation on Installing Python Modules.

Build Requirements#

To build this software, you need Python and a C++ compiler installed on your system. The C++ compiler must support the ISO C++ 2017 standard.

Additionally, the following development libraries are required:

On Ubuntu, you can install these dependencies using the following command:

sudo apt-get install libboost-date-time-dev libboost-python-dev \
    libboost-regex-dev libboost-thread-dev libnetcdf-dev \
    libudunits2-dev

Build Instructions#

Once all requirements are installed, you can build the library by running the following command at the root of the project:

python3 setup.py build

You can customize the build process by specifying additional options:

  • --boost-root: Specify the Boost installation prefix.

  • --cxx-compiler: Specify the C++ compiler to use.

  • --netcdf-dir: Specify the NetCDF installation prefix.

  • --reconfigure: Force CMake to reconfigure the project.

  • --udunits2-root: Specify the UDUNITS-2 installation prefix.

python3 setup.py build_ext --inplace --boost-root=/path/to/boost \
    --cxx-compiler=g++ --netcdf-dir=/path/to/netcdf --reconfigure \
    --udunits2-root=/path/to/udunits2

To view all available build options, run:

python3 setup.py build_ext --help

Testing#

Requirements#

Running tests require the following Python libraries:

Running tests#

The distribution contains a set of test cases that can be processed with the standard Python test framework. To run the full test suite, use the following at the root of the project:

pytest

Install#

To install the package, use pip . Run the following command at the root of the project:

pip install .

You can specify an alternate installation path with the --prefix option:

pip install . --prefix=/usr/local
Contents