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

.. only:: html

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

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

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

.. _sphx_glr_generated_examples_classification_plot_classifier_masks.py:


Classification masks
--------------------

:doc:`Classification algorithms <../../../algorithms/classification>` estimate the baseline
by classifying each point as belonging to either the baseline or the peaks. When
first using a function, the correct parameters may not be known. To make the effects
of input parameters on the classification process more easily understood, all functions
in the classification module provide a 'mask' item in the output parameter dictionary.
The mask parameter is a Boolean numpy array that is True for any point classified as
belonging to the baseline and False otherwise.

.. GENERATED FROM PYTHON SOURCE LINES 15-41

.. code-block:: Python


    import matplotlib.pyplot as plt
    import numpy as np

    from pybaselines import Baseline
    from pybaselines.utils import gaussian


    half_window = 50
    x = np.linspace(0, 1000, 1000)
    signal = (
        gaussian(x, 9, 100, 12)
        + gaussian(x, 6, 180, 5)
        + gaussian(x, 8, 350, 11)
        + gaussian(x, 15, 400, 18)
        + gaussian(x, 6, 550, 6)
        + gaussian(x, 13, 700, 8)
        + gaussian(x, 9, 800, 9)
        + gaussian(x, 9, 880, 7)
    )
    baseline = gaussian(x, -6, 700, 500)
    noise = np.random.default_rng(0).normal(0, 0.1, x.size)
    y = signal + baseline + noise

    baseline_fitter = Baseline(x_data=x)








.. GENERATED FROM PYTHON SOURCE LINES 43-47

When first fitting a new dataset, it may be difficult to estimate the correct
parameters. For this example, the main parameter for the baseline function is
the `half_window` used for the rolling standard deviation calculation. Try a low
and high value to see the difference.

.. GENERATED FROM PYTHON SOURCE LINES 47-58

.. code-block:: Python


    half_window_1 = 15
    half_window_2 = 45
    fit_1, params_1 = baseline_fitter.std_distribution(y, half_window_1, smooth_half_window=10)
    fit_2, params_2 = baseline_fitter.std_distribution(y, half_window_2, smooth_half_window=10)

    plt.plot(x, y)
    plt.plot(x, fit_1, label=f'half_window={half_window_1}')
    plt.plot(x, fit_2, '--', label=f'half_window={half_window_2}')
    plt.legend()




.. image-sg:: /generated/examples/classification/images/sphx_glr_plot_classifier_masks_001.png
   :alt: plot classifier masks
   :srcset: /generated/examples/classification/images/sphx_glr_plot_classifier_masks_001.png
   :class: sphx-glr-single-img





.. GENERATED FROM PYTHON SOURCE LINES 59-62

The two baselines are similar in most regions except for the two small peaks.
To investigate why such different results were obtained, the `mask` item in
the output parameter dictionary can be used.

.. GENERATED FROM PYTHON SOURCE LINES 62-74

.. code-block:: Python

    mask_1 = params_1['mask']
    mask_2 = params_2['mask']

    _, (ax, ax2) = plt.subplots(2, sharex=True, gridspec_kw={'hspace': 0})
    ax.plot(x, y)
    patch_1 = ax.plot(x[mask_1], y[mask_1], 'o')[0]
    ax2.plot(x, y)
    patch_2 = ax2.plot(x[mask_2], y[mask_2], 'ms')[0]
    ax.legend((patch_1, patch_2), (f'half_window={half_window_1}', f'half_window={half_window_2}'))

    plt.show()




.. image-sg:: /generated/examples/classification/images/sphx_glr_plot_classifier_masks_002.png
   :alt: plot classifier masks
   :srcset: /generated/examples/classification/images/sphx_glr_plot_classifier_masks_002.png
   :class: sphx-glr-single-img





.. GENERATED FROM PYTHON SOURCE LINES 75-78

After comparing the two masks, it is clear that the higher half_window value
mis-identified the small peaks as belonging to the baseline. Thus, the smaller
half_window is the better parameter.


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

   **Total running time of the script:** (0 minutes 0.109 seconds)


.. _sphx_glr_download_generated_examples_classification_plot_classifier_masks.py:

.. only:: html

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

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

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

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

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

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

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


.. only:: html

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

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