.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/ssa/plot_01_ssa_basics.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_ssa_plot_01_ssa_basics.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_ssa_plot_01_ssa_basics.py:


Basic SSA decomposition and frequency-guided cleaning
=====================================================

This network-free example separates a slow synthetic drift from an alpha
rhythm. The frequency grouping is an mne-denoise extension, not a grouping rule
prescribed by Basic SSA.

.. GENERATED FROM PYTHON SOURCE LINES 9-54



.. rst-class:: sphx-glr-horizontal


    *

      .. image-sg:: /auto_examples/ssa/images/sphx_glr_plot_01_ssa_basics_001.png
         :alt: Frequency-guided Basic SSA
         :srcset: /auto_examples/ssa/images/sphx_glr_plot_01_ssa_basics_001.png
         :class: sphx-glr-multi-img

    *

      .. image-sg:: /auto_examples/ssa/images/sphx_glr_plot_01_ssa_basics_002.png
         :alt: PSD Comparison
         :srcset: /auto_examples/ssa/images/sphx_glr_plot_01_ssa_basics_002.png
         :class: sphx-glr-multi-img


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

 .. code-block:: none

    Additive reconstruction error: 2.520e-14
    Dropped frequencies: [0.5 0.5]

    <Figure size 1600x800 with 1 Axes>





|

.. code-block:: Python


    import numpy as np

    from mne_denoise.ssa import SingularSpectrumAnalysis, ssa_decompose
    from mne_denoise.viz import plot_psd_comparison, plot_signal_overlay

    sfreq = 200.0
    times = np.arange(1200) / sfreq
    alpha = np.sin(2 * np.pi * 10.0 * times)
    drift = 4.0 * np.sin(2 * np.pi * 0.5 * times)
    rng = np.random.default_rng(4)
    observed = drift + alpha + 0.05 * rng.standard_normal(times.size)

    components, info = ssa_decompose(observed, window_seconds=0.5, sfreq=sfreq)
    reconstruction_error = np.max(np.abs(components.sum(axis=0) - observed))
    print(f"Additive reconstruction error: {reconstruction_error:.3e}")

    cleaner = SingularSpectrumAnalysis(
        sfreq=sfreq,
        window_seconds=0.5,
        drop_freq_max=3.0,
    )
    cleaned = cleaner.fit_transform(observed[np.newaxis])[0]
    print(f"Dropped frequencies: {cleaner.dropped_frequencies_[0]}")

    plot_signal_overlay(
        observed,
        cleaned,
        times,
        reference=alpha,
        before_label="Observed",
        after_label="Basic SSA",
        reference_label="Alpha reference",
        scale_after=False,
        title="Frequency-guided Basic SSA",
        show=False,
    )
    plot_psd_comparison(
        observed,
        cleaned,
        sfreq=sfreq,
        fmin=0,
        fmax=25,
        show=False,
    )


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

   **Total running time of the script:** (0 minutes 0.581 seconds)


.. _sphx_glr_download_auto_examples_ssa_plot_01_ssa_basics.py:

.. only:: html

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

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

      :download:`Download Jupyter notebook: plot_01_ssa_basics.ipynb <plot_01_ssa_basics.ipynb>`

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

      :download:`Download Python source code: plot_01_ssa_basics.py <plot_01_ssa_basics.py>`

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

      :download:`Download zipped: plot_01_ssa_basics.zip <plot_01_ssa_basics.zip>`