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

.. only:: html

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

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

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

.. _sphx_glr_examples_extracting_data.py:


Extracting Data from Nodes
---------------
In this example we overlay a resulting mapper graph on top of
the scatter-plot of datapoints. The nodes will be placed at
the mean of the corresponding cluster. The color of the node
will correspond to the standard deviation.

.. GENERATED FROM PYTHON SOURCE LINES 11-13

Generate Data
=============

.. GENERATED FROM PYTHON SOURCE LINES 13-35

.. code-block:: Python

    import matplotlib.pyplot as plt
    import numpy as np

    # For reproducibility
    np.random.set_state(np.random.RandomState().get_state(137))


    def generate_circle(n_points=100, radius=1, center_x=0, center_y=0):
        theta = np.random.uniform(0, 2 * np.pi, n_points)
        x = radius * np.cos(theta) + center_x
        y = radius * np.sin(theta) + center_y
        return np.c_[x, y]


    circle1 = generate_circle()
    circle2 = generate_circle(400, 2, 3, 0)
    circle3 = generate_circle(100, 1, 3, 0)
    data = np.vstack((circle1, circle2, circle3))

    plt.scatter(data[:, 0], data[:, 1])
    plt.show()




.. image-sg:: /examples/images/sphx_glr_extracting_data_001.png
   :alt: extracting data
   :srcset: /examples/images/sphx_glr_extracting_data_001.png
   :class: sphx-glr-single-img





.. GENERATED FROM PYTHON SOURCE LINES 36-38

Implement Mapper
================

.. GENERATED FROM PYTHON SOURCE LINES 38-53

.. code-block:: Python

    import zen_mapper as zm
    from sklearn.cluster import DBSCAN

    cover_scheme = zm.Width_Balanced_Cover(n_elements=7, percent_overlap=0.2)
    projection = data[:, 0]
    clusterer = zm.sk_learn(DBSCAN(eps=0.5))

    result = zm.mapper(
        data=data,
        projection=projection,
        clusterer=clusterer,
        cover_scheme=cover_scheme,
        dim=1,
    )








.. GENERATED FROM PYTHON SOURCE LINES 54-60

Accessing Data Points from Nodes
================================
The ``MapperResult.nodes`` attribute holds a list of clusters which are
each a list of data indices belonging to the cluster.
Once you have access to these indices you can run your
own analysis.

.. GENERATED FROM PYTHON SOURCE LINES 60-76

.. code-block:: Python


    cluster_centers = []
    cluster_stds = []

    # Get data from nodes!
    for data_indices in result.nodes:
        # This is where we extract data
        cluster_points = data[data_indices]
        center = np.mean(cluster_points, axis=0)
        std = np.std(cluster_points)
        cluster_centers.append(center)
        cluster_stds.append(std)

    node_positions = np.array(cluster_centers)
    node_stds = np.array(cluster_stds)








.. GENERATED FROM PYTHON SOURCE LINES 77-82

Visualize
====================================
We can access all k-simplices by calling the ``MapperResult.nerve[k]`` attribute. This will return simplices stored as a (k+1)-tuple
of node_ids. One can then apply the same process as above to access the corresponding data points for analysis/visualization.
For now, we will just use this to draw edges manually.

.. GENERATED FROM PYTHON SOURCE LINES 82-106

.. code-block:: Python


    plt.scatter(data[:, 0], data[:, 1], alpha=0.1, c="gray", s=20)

    # draw edges from nerve attribute
    for node1_id, node2_id in result.nerve[1]:
        x_coords = [node_positions[node1_id, 0], node_positions[node2_id, 0]]
        y_coords = [node_positions[node1_id, 1], node_positions[node2_id, 1]]
        plt.plot(x_coords, y_coords, "k-", alpha=0.6, linewidth=2)

    # draw nodes
    scatter = plt.scatter(
        node_positions[:, 0],
        node_positions[:, 1],
        c=node_stds,
        s=100,
        cmap="viridis",
        edgecolors="black",
        linewidth=1,
        zorder=2,  # just places nodes above edges
    )

    plt.colorbar(scatter, label="Cluster \n Standard Deviation")
    plt.title("Decorated Mapper Graph")
    plt.show()



.. image-sg:: /examples/images/sphx_glr_extracting_data_002.png
   :alt: Decorated Mapper Graph
   :srcset: /examples/images/sphx_glr_extracting_data_002.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** (0 minutes 0.177 seconds)


.. _sphx_glr_download_examples_extracting_data.py:

.. only:: html

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

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

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

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

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

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

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


.. only:: html

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

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