Setup and Installation

Dependencies

IceAdvect is dependent on several open source programs that can be installed using OS-specific package management systems (e.g. apt or homebrew), conda or from source:

• PROJ

• HDF5

• libxml2

• libxslt

Installation

IceAdvect is available for download from the GitHub repository, the Python Package Index (pypi), and from conda-forge.

The simplest installation for most users will likely be using conda or mamba:

conda install -c conda-forge iceadvect

conda installed versions of IceAdvect can be upgraded to the latest stable release:

conda update iceadvect

Development Install

To use the development repository, please fork IceAdvect into your own account and then clone onto your system:

git clone https://github.com/tsutterley/IceAdvect.git

IceAdvect can then be installed within the package directory using pip:

python3 -m pip install --user .

To include all optional dependencies:

python3 -m pip install --user .[all]

The development version of IceAdvect can also be installed directly from GitHub using pip:

python3 -m pip install --user git+https://github.com/tsutterley/IceAdvect.git

Package Management with pixi

Alternatively pixi can be used to create a streamlined environment after cloning the repository:

pixi install

pixi maintains isolated environments for each project, allowing for different versions of IceAdvect and its dependencies to be used without conflict. The pixi.lock file within the repository defines the required packages and versions for the environment.

pixi can also create shells for running programs within the environment:

pixi shell

To see the available tasks within the IceAdvect workspace:

pixi task list

Tasks that can run on this machine:
-----------------------------------
coverage (by design), docs (by design), export-env (by design), lint (by design), start (by design), test (by design)
Task        Description
coverage    Run the pytest suite and create coverage reports
docs        Create HTML documentation with sphinx
export-env  Export workspace to a conda environment file
lint        Run flake8 to check for compilation errors
start       Start JupyterLab in the notebooks directory
test        Run the pytest suite

Note

pixi is under active development and may change in future releases

Verifying Installation

import IceAdvect

print(f"IceAdvect version: {IceAdvect.__version__}")

IceAdvect version: 0.0.1