.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/asr/plot_01_asr_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_asr_plot_01_asr_basics.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_asr_plot_01_asr_basics.py:


Artifact Subspace Reconstruction: Basic Usage.
==============================================

This example demonstrates standard ASR on a synthetic multichannel EEG signal
with short spatial burst artifacts.

.. GENERATED FROM PYTHON SOURCE LINES 10-12

Imports
-------

.. GENERATED FROM PYTHON SOURCE LINES 12-19

.. code-block:: Python

    import matplotlib.pyplot as plt
    import numpy as np

    from mne_denoise.asr import ASR
    from mne_denoise.qa import variance_removed
    from mne_denoise.viz import plot_signal_overlay








.. GENERATED FROM PYTHON SOURCE LINES 20-22

Synthetic EEG with Bursts
-------------------------

.. GENERATED FROM PYTHON SOURCE LINES 22-49

.. code-block:: Python

    sfreq = 250.0
    duration = 12.0
    n_channels = 8
    n_times = int(sfreq * duration)
    times = np.arange(n_times) / sfreq
    rng = np.random.default_rng(42)

    brain = np.zeros((n_channels, n_times))
    for ch_idx in range(n_channels):
        phase = rng.uniform(0, 2 * np.pi)
        brain[ch_idx] = (
            0.5 * np.sin(2 * np.pi * 10 * times + phase)
            + 0.2 * np.sin(2 * np.pi * 6 * times + 0.5 * phase)
            + 0.05 * rng.standard_normal(n_times)
        )

    data = brain.copy()
    burst_mask = np.zeros(n_times, dtype=bool)
    spatial = rng.standard_normal((n_channels, 2))
    spatial /= np.linalg.norm(spatial, axis=0, keepdims=True)

    for onset, stop in ((4.0, 4.8), (8.0, 8.6)):
        start = int(onset * sfreq)
        end = int(stop * sfreq)
        burst_mask[start:end] = True
        data[:, start:end] += spatial @ (8.0 * rng.standard_normal((2, end - start)))








.. GENERATED FROM PYTHON SOURCE LINES 50-55

Fit and Apply ASR
-----------------
The value 20 is retained for comparison with common reference workflows.
It is not portable across calibration rules or implementations; tune and freeze
an operating point using artifact-attenuation and neural-preservation endpoints.

.. GENERATED FROM PYTHON SOURCE LINES 55-69

.. code-block:: Python

    asr = ASR(
        sfreq=sfreq,
        cutoff=20.0,
        calibration="auto",
        filter_kind="none",
        max_dims=0.5,
    )
    clean = asr.fit_transform(data)

    print(f"Calibration windows kept: {asr.clean_window_mask_.sum()}")
    print(f"Repaired windows: {(asr.n_components_reconstructed_ > 0).sum()}")
    print(f"Repaired sample fraction: {asr.fraction_reconstructed_samples_:.2%}")
    print(f"Variance removed: {variance_removed(data, clean):.2f}%")





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

 .. code-block:: none

    Calibration windows kept: 22
    Repaired windows: 11
    Repaired sample fraction: 25.63%
    Variance removed: 92.87%




.. GENERATED FROM PYTHON SOURCE LINES 70-75

Plot One Channel
----------------
``plot_signal_overlay`` draws the before/after pair and, via its optional
``reference`` and ``highlight_mask`` arguments, overlays the ground-truth
signal and shades the burst windows directly.

.. GENERATED FROM PYTHON SOURCE LINES 75-94

.. code-block:: Python

    plot_signal_overlay(
        data,
        clean,
        times,
        pick=0,
        scale_after=False,
        before_label="Noisy",
        after_label="ASR cleaned",
        x_label="Time (s)",
        y_label="Amplitude (a.u.)",
        title="ASR burst repair",
        reference=brain[0],
        reference_label="Reference signal",
        highlight_mask=burst_mask,
        highlight_label="Burst artifact",
        show=False,
    )

    plt.show()



.. image-sg:: /auto_examples/asr/images/sphx_glr_plot_01_asr_basics_001.png
   :alt: ASR burst repair
   :srcset: /auto_examples/asr/images/sphx_glr_plot_01_asr_basics_001.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** (0 minutes 0.633 seconds)


.. _sphx_glr_download_auto_examples_asr_plot_01_asr_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_asr_basics.ipynb <plot_01_asr_basics.ipynb>`

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

      :download:`Download Python source code: plot_01_asr_basics.py <plot_01_asr_basics.py>`

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

      :download:`Download zipped: plot_01_asr_basics.zip <plot_01_asr_basics.zip>`