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

.. only:: html

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

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

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

.. _sphx_glr_auto_examples_plot_moon_clustering.py:


================================================================================================
Drawing a decision boundary between two interlacing moons
================================================================================================

This example is a retake from the experiment in the original GEMINI paper where we want to find the true clusters
in between two facing moons. To do so, the trick is to use a specific distance using the "precomputed" option
which will guide the clustering algorithm to the desired solution.

Note that we use `gemclus.mlp.MLPWasserstein` because a linear model would not be able to find the optimal boundary.

.. GENERATED FROM PYTHON SOURCE LINES 12-20

.. code-block:: default


    import numpy as np
    from matplotlib import pyplot as plt
    from scipy.sparse import csgraph
    from sklearn import datasets, metrics

    from gemclus.mlp import MLPWasserstein








.. GENERATED FROM PYTHON SOURCE LINES 21-23

Generate two interlacing moons
--------------------------------------------------------------

.. GENERATED FROM PYTHON SOURCE LINES 25-27

.. code-block:: default

    X, y = datasets.make_moons(n_samples=200, noise=0.05, random_state=2023)








.. GENERATED FROM PYTHON SOURCE LINES 28-30

Pre-compute a specific metric between samples
--------------------------------------------------------------

.. GENERATED FROM PYTHON SOURCE LINES 32-45

.. code-block:: default


    # Create an adjacency graph where edges are defined if the distance between two samples is
    # less than the 5% quantile of the Euclidean distances
    distances = metrics.pairwise_distances(X, metric="euclidean")
    threshold = np.quantile(distances, 0.05)
    adjacency = distances < threshold

    # compute the all-pairs shortest path in this graph
    distances = csgraph.floyd_warshall(adjacency, directed=False, unweighted=True)

    # Replace np.inf with 2 times the size of the matrix
    distances[np.isinf(distances)] = 2 * distances.shape[0]








.. GENERATED FROM PYTHON SOURCE LINES 46-49

Train the model
--------------------------------------------------------------
Note that we use the precomputed option and pass our distance to the `.fit` function along `X`.

.. GENERATED FROM PYTHON SOURCE LINES 51-54

.. code-block:: default

    model = MLPWasserstein(n_clusters=2, metric="precomputed", random_state=2023, learning_rate=1e-2)
    y_pred = model.fit_predict(X, distances)








.. GENERATED FROM PYTHON SOURCE LINES 55-57

Final Clustering
-----------------

.. GENERATED FROM PYTHON SOURCE LINES 59-70

.. code-block:: default


    x_vals = np.linspace(X[:, 0].min() - 0.5, X[:, 0].max() + 0.5, num=50)
    y_vals = np.linspace(X[:, 1].min() - 0.5, X[:, 1].max() + 0.5, num=50)
    xx, yy = np.meshgrid(x_vals, y_vals)
    grid_inputs = np.c_[xx.ravel(), yy.ravel()]
    zz = model.predict(grid_inputs).reshape((50, 50))

    plt.contourf(xx, yy, zz, alpha=0.3, cmap=plt.cm.Spectral)
    plt.scatter(X[:, 0], X[:, 1], c=y_pred)
    plt.axis("off")
    plt.show()



.. image-sg:: /auto_examples/images/sphx_glr_plot_moon_clustering_001.png
   :alt: plot moon clustering
   :srcset: /auto_examples/images/sphx_glr_plot_moon_clustering_001.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** (0 minutes 7.876 seconds)


.. _sphx_glr_download_auto_examples_plot_moon_clustering.py:

.. only:: html

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




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

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

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

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


.. only:: html

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

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