The goal of the library is to provide base functions of Rough Sets Theory and give foundation to build extensions which will provide different methods built on RoughSets Theory, like:
- pre-processing methods
- find core and reducts
- classifiers
- post-processing methods
The library doesn't use basic loops so should help to build very fast extensions when large datasets will be used.
The library implements these main functions:
-
computation of indiscernibilty relations - function: get_indiscernibility_relations
-
computation of a lower and upper approximations, boundary and negative regions - all these 4 boundaries are computed by function: get_approximation_indices. For optimization, only indices of X,y are returned by the function, so can be used for futher computations
before slicing with X and y.
The library has included unit tests for different datasets, subsets and concepts.
Python >= 3.8
OS: Linux, Windows
pip install roughsets-base
pip install --upgrade pip
pip install --upgrade build
On linux: python3 -m build
On Windows:
py -m build
pip install dist/roughsets_base--py3-none-any.whl
pip install -r requirements.dev.txt
File tests/test_dataset_KDD.py contains tests which use KDD99 dataset.
By default file will be downloaded from: http://kdd.ics.uci.edu/databases/kddcup99/corrected.gz
If You have the file on Your system You can set OS environment variable ROUGHSETS_KDD99_TEST_DATA_FOLDER
with the path to th file.
File tests/KDD99_compare_with_R_RoughSets.R contains R script which generate refernece data using well knwon reference library: RoughSets.
See: https://www.rdocumentation.org/packages/RoughSets/topics/RoughSets-package (R language).
Reference datasets with results from R-RoughSets library are saved in folder tests/datasets/KDD99.
You can disable running tests for specific dataset in file test_dataset_X.py (X - symbol of dataset), method setUp.
For checking of unit tests' results was used algorithms from well known reference library:
https://www.rdocumentation.org/packages/RoughSets/topics/RoughSets-package
pip install -r requirements.dev.txt
sphinx-build -b html ./doc ./doc/_build/html
sphinx-build -b man ./doc ./doc/_build/man
sklearn-pandas https://github.com/scikit-learn-contrib/sklearn-pandas
pandas ecosystem: https://pandas.pydata.org/community/ecosystem.html
If You would like to use this library in Your research, please cite it.
Jankowski, D. (2022). Roughsets-base: Python library for base Roughsets methods. (v1.0.1) [Computer software]. Zenodo. https://doi.org/10.5281/ZENODO.5957474
