.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "examples/preprocessors/frps.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

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

        :ref:`Go to the end <sphx_glr_download_examples_preprocessors_frps.py>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr_examples_preprocessors_frps.py:


============================
Instance selection with FRPS
============================

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.

.. GENERATED FROM PYTHON SOURCE LINES 16-81



.. image-sg:: /examples/preprocessors/images/sphx_glr_frps_001.png
   :alt: FRNN applied to instances of iris dataset selected by FRPS, upper, lower, both
   :srcset: /examples/preprocessors/images/sphx_glr_frps_001.png
   :class: sphx-glr-single-img





.. code-block:: Python

    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.classifiers import FRNN
    from frlearn.instance_preprocessors import FRPS
    from frlearn.t_norms import heyting_t_norm, lukasiewicz_t_norm

    # 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_t_norm), ('Heyting', heyting_t_norm)]):
        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(X_orig, y_orig)

            # Create an instance of the FRNN classifier and construct the model.
            clf = FRNN(upper_weights=None, lower_weights=None, upper_k=1, lower_k=1)
            model = clf(X, y)

            # Query mesh points to obtain class values and select highest valued class.
            Z = model(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.get_subplotspec().is_first_col():
                plt.ylabel(aggr_name, rotation=0, size='large', ha='right')
            if axes.get_subplotspec().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 2.664 seconds)


.. _sphx_glr_download_examples_preprocessors_frps.py:

.. only:: html

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

    .. container:: sphx-glr-download sphx-glr-download-jupyter

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

    .. container:: sphx-glr-download sphx-glr-download-python

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

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: frps.zip <frps.zip>`


.. only:: html

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

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