.. note::
    :class: sphx-glr-download-link-note

    Click :ref:`here <sphx_glr_download_auto_examples_neighbours_plot_frps_preprocessing.py>` to download the full example code
.. rst-class:: sphx-glr-example-title

.. _sphx_glr_auto_examples_neighbours_plot_frps_preprocessing.py:


==================
FRPS preprocessing
==================

Sample usage of FRPS preprocessing, demonstrated in combination with (strict) FRNN classification.

The figures contain the selected prototypes within a section of the feature space.
The prototypes are coloured according to their true labels,
while the feature space is coloured according to predictions on the basis of the prototypes,
making the decision boundaries visible.

In total, nine subfigures are displayed,
to illustrate the effect of the `quality_measure` (rows) and `aggr_R` (columns) parameters.



.. image:: /auto_examples/neighbours/images/sphx_glr_plot_frps_preprocessing_001.png
    :class: sphx-glr-single-img


.. rst-class:: sphx-glr-script-out

 Out:

 .. code-block:: none


    /home/oliver/code/scikit-learn-contrib/fuzzy-rough-learn/examples/neighbours/plot_frps_preprocessing.py:80: UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure.
      plt.show()





|


.. code-block:: default

    print(__doc__)

    import numpy as np
    import matplotlib.pyplot as plt
    from matplotlib.colors import ListedColormap
    from sklearn import datasets

    from frlearn.base import select_class
    from frlearn.neighbours import FRNN, FRPS
    from frlearn.utils.owa_operators import strict
    from frlearn.utils.t_norms import lukasiewicz

    # Import example data and reduce to 2 dimensions.
    iris = datasets.load_iris()
    X_orig = iris.data[:, :2]
    y_orig = iris.target

    # Create a mesh of points in the attribute space.
    step_size = .02
    x_min, x_max = X_orig[:, 0].min() - 1, X_orig[:, 0].max() + 1
    y_min, y_max = X_orig[:, 1].min() - 1, X_orig[:, 1].max() + 1
    xx, yy = np.meshgrid(np.arange(x_min, x_max, step_size), np.arange(y_min, y_max, step_size))

    # Define color maps.
    cmap_light = ListedColormap(['#FFAAAA', '#AAFFAA', '#AAAAFF'])
    cmap_bold = ListedColormap(['#FF0000', '#00FF00', '#0000FF'])

    # Initialise figure.
    plt.figure()

    for i, (aggr_name, aggr_R) in enumerate([('mean', np.mean), ('Łukasiewicz', lukasiewicz), ('min', np.amin)]):
        for j, quality_measure in enumerate(['upper', 'lower', 'both']):
            axes = plt.subplot(3, 3, i*3 + j + 1)

            # Create an instance of the FRPS preprocessor and process the data.
            preprocessor = FRPS(aggr_R=aggr_R, quality_measure=quality_measure)
            X, y = preprocessor.process(X_orig, y_orig)

            # Create an instance of the FRNN classifier and construct the model.
            clf = FRNN(upper_weights=strict(), lower_weights=strict(), upper_k=1, lower_k=1)
            model = clf.construct(X, y)

            # Query mesh points to obtain class values and select highest valued class.
            Z = model.query(np.c_[xx.ravel(), yy.ravel()])
            Z = select_class(Z, labels=model.classes)

            # Plot mesh.
            Z = Z.reshape(xx.shape)
            plt.pcolormesh(xx, yy, Z, cmap=cmap_light)

            # Plot training instances.
            plt.scatter(X[:, 0], X[:, 1], c=y, cmap=cmap_bold, edgecolor='k', s=20)

            # Set plot dimensions.
            plt.xlim(xx.min(), xx.max())
            plt.ylim(yy.min(), yy.max())

            # Describe columns and rows.
            if axes.is_first_col():
                plt.ylabel(aggr_name, rotation=0, size='large', ha='right')
            if axes.is_first_row():
                plt.title(quality_measure)

    plt.suptitle('FRNN applied to instances of iris dataset selected by FRPS', fontsize=14)
    plt.show()


.. rst-class:: sphx-glr-timing

   **Total running time of the script:** ( 0 minutes  4.547 seconds)


.. _sphx_glr_download_auto_examples_neighbours_plot_frps_preprocessing.py:


.. only :: html

 .. container:: sphx-glr-footer
    :class: sphx-glr-footer-example



  .. container:: sphx-glr-download

     :download:`Download Python source code: plot_frps_preprocessing.py <plot_frps_preprocessing.py>`



  .. container:: sphx-glr-download

     :download:`Download Jupyter notebook: plot_frps_preprocessing.ipynb <plot_frps_preprocessing.ipynb>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_
