Installation

Create and environment

To streamline the use of PyForTraCC and manage dependencies effectively, it’s recommended to create a dedicated environment named pyfortracc. This environment can be created with Anaconda or by using a pip environment, depending on your preferred setup.

Option 1: Using Anaconda or Miniconda

Anaconda and Miniconda are popular tools for managing data science environments and packages. They provide an isolated setup that ensures compatibility between dependencies.

• Create the environment from the environment.yml file:

conda env create -f environment.yml

• Activate the environment:

conda activate pyfortracc

Option 2: Using pip and a Virtual Environment

If you prefer using pip, you can create a virtual environment using venv and install the required packages.

• Create a virtual environment:

python3 -m venv pyfortracc

• Activate the environment in macOS and Linux:

source pyfortracc/bin/activate

• Activate the environment in Windows:

pyfortracc\Scripts\activate

Install

To install PyForTraCC, we recommend using one of the following methods based on your needs. The primary method listed ensures that you are using the latest version directly from the official repository, with alternative methods available for convenience.

Recommended Method: Install Latest Version via GitHub

For the most up-to-date version of PyForTraCC, we recommend installing directly from the GitHub repository. This method allows you to obtain any recent updates or bug fixes made to the package:

pip3 install --upgrade git+https://github.com/fortracc/pyfortracc.git@main#egg=pyfortracc

This command installs PyForTraCC from the main branch, ensuring you have the latest stable version available.

Alternative Method: Install via Conda

If you prefer to use conda for package management, PyForTraCC is also available through the conda-forge channel. While this version might lag slightly behind the latest GitHub version, it is a convenient option for users within the conda ecosystem:

conda install -c conda-forge pyfortracc

Alternative Method: Install via PyPI

PyForTraCC is available on PyPI as well. This option provides a stable release but may not include the latest features or fixes. To install from PyPI, use:

pip3 install pyfortracc

Each of these methods will install PyForTraCC, but we strongly recommend the GitHub installation if you want the most current version for ongoing developments and improvements.

List of requirements

The following libraries are required to ensure that PyForTraCC functions optimally. These dependencies support various essential tasks such as data handling, geographic and spatial operations, scientific computing, and performance monitoring. Before running the PyForTraCC library, please ensure these packages are installed within your environment.

• rasterio

• geopandas

• opencv-python

• opencv_contrib_python

• xarray

• scipy

• scikit-learn

• pyarrow

• netCDF4

• cartopy

• tqdm

• ipython

• ipykernel

• psutil

• duckdb

• shapelysmooth

We recommend installing the pyForTraCC package with Python 3.12. To simplify the installation process and avoid potential conflicts, it’s advisable to use a virtual environment (e.g., Anaconda3, Miniconda, Mamba) for managing dependencies.