py-mltoolkit

py-mltoolkit is a set of utilities for machine learning projects in Python

• Requirements
• Installation
• Usage
• Development
  • Code style
  • Testing
  • Document
• License

Requirements

The package has been written in Python 3, and it needs version 3.11.

The dependencies are managed by pip using the file requirements.txt.

Installation

To add py-mltoolkit to your project, run the following command. This will add the last released version to you project:

pip install --upgrade cerbernetix.mltoolkit

For getting the last unstable version, the package needs to be installed from the source code. The following command will grab the last develop version:

pip install --upgrade git+https://github.com/cerbernetix/py-mltoolkit.git@develop

If you prefer using ssh:

pip install --upgrade git+ssh://[email protected]/cerbernetix/py-mltoolkit.git@develop

Usage

py-mltoolkit offers several utilities per domain.

Please refer to the documentation for more information.

Development

Check out the repository:

git clone [email protected]:cerbernetix/py-mltoolkit.git

Then, create the virtual env and install the dependencies:

cd py-mltoolkit
python3 -m venv ".venv"
source ".venv/bin/activate"
pip install -r requirements.txt
pip install -e .

Note: For deactivating the virtual env, call the command deactivate.

Automating the environment activation/deactivation

For activating the virtual env automatically when entering the project folder, and deactivating it when leaving the folder, you can add this snippet to you shell profile:

cd() {
    builtin cd "$@"

    local venv=".venv"

    # If a Python virtualenv is active, deactivate it if the new folder is outside
    if [[ -v VIRTUAL_ENV ]] ; then
        local parent=$(dirname "${VIRTUAL_ENV}")
        if [[ "${PWD}"/ != "${parent}"/* ]] ; then
            deactivate
        fi
    fi

    # If a Python env folder is found then activate the virtualenv
    if [[ -d "./${venv}" ]] ; then
        # Is it a Python venv?
        if [[ -f "./${venv}/bin/activate" ]] ; then
            source "./${venv}/bin/activate"
        fi
    fi
}

Code style

Code is linted using PyLint and formatted using Black. The docstrings are validated using pydocstyle.

Testing

Each module comes with unit tests, by convention, a test folder must be added to each package

Unit tests are made using unittest. To run them:

python3 -m unittest

Document

Docstrings are written following the Google docstrings format.

The documentation is generated using lazydocs.

The script ./pydoc.sh can be used to regenerate it.

License

Copyright (c) 2024 Jean-Sébastien CONAN Distributed under the MIT License (See LICENSE file or copy at http://opensource.org/licenses/MIT).