Tests
Package
Metadata

subgroups is a public, accessible and open-source python library created to work with the Subgroup Discovery (SD) technique. This library implements the necessary components related to the SD technique and contains a collection of SD algorithms and other data analysis utilities.

The easiest way to obtain this library is from either PyPI (the Python Package Index) or Conda.

For that, you can view and download the package from its PyPI page or directly install it by executing:

pip install subgroups

For that, you can view and download the package from its Anaconda.org page (conda-forge channel) or directly install it by executing:

conda install -c conda-forge subgroups

After installing the library, a collection of tests can be launched by executing:

import subgroups.tests as st
st.run_all_tests()

These tests verify that the library is correctly installed and that all components, algorithms and features are properly working.

The source code (latest development) is currently hosted on: https://github.com/antoniolopezmc/subgroups

Therefore, you need first to clone the repository:

git clone https://github.com/antoniolopezmc/subgroups.git
cd subgroups

After that, the library can be installed in production mode or in develop mode.

make install_prod

or

python -m pip install ./

or

pip install ./

This mode installs the library as normal, copying it to the standard Python site-packages directory.

make install_dev

or

python -m pip install -e ./

or

pip install -e ./

This mode installs the library in editable mode, creating a link in the standard Python site-packages directory to the downloaded project directory (the current directory). See the pip_install documentation for further details.

An example of use of each algorithm implemented in subgroups python library can be found in the examples/algorithms folder:

• SDMap
• SDMap*
• VLSD
• BSD-CBSD-CPBSD
• QFinder
• GMSL
• DSLM

The official documentation is hosted on https://www.um.es/subgroups/

Additionally, the source code of the project contains a folder called docs, which includes the documentation of the library. This documentation can be also manually generated by executing:

cd docs
make build

or

cd docs
python clean.py source/project_files build
python -m pip install --upgrade sphinx
python -m pip install --upgrade sphinx-rtd-theme
python -m pip install --upgrade sphinx-autodoc-typehints
sphinx-apidoc -f -T -M -o source/project_files ../src/subgroups
sphinx-build -M html source build

The generated documentation will be located in the build subfolder.