.. -- mode: rst --
.. _scikit-learn: https://scikit-learn.org
.. _documentation: https://sklearn-sfa.readthedocs.io/en/latest/index.html
.. _MDP: https://mdp-toolkit.github.io/
.. _PyPi: https://pypi.org/project/sklearn-sfa/
sklearn-sfa or sksfa is an implementation of Slow Feature Analysis for scikit-learn_.
It is meant as a standalone transformer for dimensionality reduction or as a building block for more complex representation learning pipelines utilizing scikit-learn's extensive collection of machine learning methods.
The package contains a solver for linear SFA and some auxiliary functions. The documentation_ provides an explanation of the algorithm, different use-cases, as well as pointers how to fully utilize SFA's potential, e.g., by employing non-linear basis functions or more sophisticated architectures.
For use with high-dimensional image data, sklearn-sfa now also includes an experimental implementation of Hierarchical SFA networks (HSFA) - please consult the introductory examples in the documentation.
Since sklearn-sfa is in its early stages, we also recommend taking a look at the Modular Toolkit for Data Processing MDP_ which provides stable SFA implementations that have stood the test of time.
The latest official version of the package can be installed from PyPi_ via pip:
.. code-block:: bash
pip install --user sklearn-sfa
To use the latest code, the package can also be cloned directly from GitHub and then be installed via:
.. code-block:: bash
cd sklearn-sfa pip install -e .
In Python 3.6+, the package can then be imported as
.. code-block:: python
import sksfa
The package comes with an SFA transformer. Below you see an example of initializing a transformer that extracts 2-dimensional features:
.. code-block:: python
sfa_transformer = sksfa.SFA(n_components=2)
The transformer implements sklearn's typical interface by providing fit, fit_transform, and transform methods.