.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/asr/plot_02_mne_raw_qc.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_02_mne_raw_qc.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_asr_plot_02_mne_raw_qc.py:


Artifact Subspace Reconstruction: Raw QC and Annotations.
=========================================================

This example shows ASR on an ``mne.io.RawArray`` with optional clean_windows-
style final window rejection. The repaired spans and rejected spans are kept
as annotations for downstream QC instead of deleting samples immediately.

.. GENERATED FROM PYTHON SOURCE LINES 11-13

Imports
-------

.. GENERATED FROM PYTHON SOURCE LINES 13-20

.. code-block:: Python

    import matplotlib.pyplot as plt
    import mne
    import numpy as np

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








.. GENERATED FROM PYTHON SOURCE LINES 21-23

Synthetic Raw
-------------

.. GENERATED FROM PYTHON SOURCE LINES 23-55

.. code-block:: Python

    sfreq = 250.0
    duration = 15.0
    n_eeg = 8
    n_times = int(sfreq * duration)
    times = np.arange(n_times) / sfreq
    rng = np.random.default_rng(7)

    brain = np.zeros((n_eeg, n_times))
    for ch_idx in range(n_eeg):
        phase = rng.uniform(0, 2 * np.pi)
        brain[ch_idx] = (
            0.4 * np.sin(2 * np.pi * 10 * times + phase)
            + 0.15 * np.sin(2 * np.pi * 6 * times + 0.5 * phase)
            + 0.04 * rng.standard_normal(n_times)
        )

    eog = 0.25 * np.sin(2 * np.pi * 1.2 * times)[np.newaxis, :]
    data = brain.copy()
    spatial = rng.standard_normal((n_eeg, 2))
    spatial /= np.linalg.norm(spatial, axis=0, keepdims=True)

    for onset, stop in ((4.5, 5.3), (9.0, 9.7), (12.0, 12.6)):
        start = int(onset * sfreq)
        end = int(stop * sfreq)
        data[:, start:end] += spatial @ (7.0 * rng.standard_normal((2, end - start)))

    raw_data = np.vstack([data, eog])
    ch_names = [f"EEG {idx:02d}" for idx in range(n_eeg)] + ["EOG 01"]
    ch_types = ["eeg"] * n_eeg + ["eog"]
    info = mne.create_info(ch_names, sfreq, ch_types)
    raw = mne.io.RawArray(raw_data, info, verbose=False)








.. GENERATED FROM PYTHON SOURCE LINES 56-58

Fit and Apply ASR
-----------------

.. GENERATED FROM PYTHON SOURCE LINES 58-77

.. code-block:: Python

    asr = ASR(
        cutoff=5.0,
        calibration="auto",
        picks="eeg",
        filter_kind="none",
        window_criterion=0.25,
        window_criterion_tolerances=(-np.inf, 2.5),
        verbose=False,
    )
    raw_clean = asr.fit_transform(raw)
    repair_annotations = asr.to_annotations()
    reject_annotations = asr.to_annotations("rejection")

    print(f"Repaired sample fraction: {asr.fraction_reconstructed_samples_:.2%}")
    print(
        "Retained after final window rejection: "
        f"{asr.fraction_retained_after_window_rejection_:.2%}"
    )





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

 .. code-block:: none

    Repaired sample fraction: 52.08%
    Retained after final window rejection: 86.67%




.. GENERATED FROM PYTHON SOURCE LINES 78-83

Plot One EEG Channel
--------------------
The repair and rejection annotations are passed straight to
``plot_signal_overlay`` as ``highlight_spans`` instead of shading the axis by
hand; the reference trace uses the ``reference`` argument.

.. GENERATED FROM PYTHON SOURCE LINES 83-124

.. code-block:: Python

    channel = "EEG 00"
    noisy = raw.get_data(picks=[channel])[0]
    clean = raw_clean.get_data(picks=[channel])[0]

    spans = [
        {
            "onset": onset,
            "duration": dur,
            "color": "C3",
            "alpha": 0.12,
            "label": "ASR repair",
        }
        for onset, dur in zip(repair_annotations.onset, repair_annotations.duration)
    ] + [
        {
            "onset": onset,
            "duration": dur,
            "color": "0.2",
            "alpha": 0.08,
            "label": "Window reject mask",
        }
        for onset, dur in zip(reject_annotations.onset, reject_annotations.duration)
    ]

    plot_signal_overlay(
        noisy,
        clean,
        times,
        scale_after=False,
        before_label="Noisy EEG",
        after_label="ASR cleaned EEG",
        x_label="Time (s)",
        y_label="Amplitude (a.u.)",
        title="ASR repairs plus optional final window rejection mask",
        reference=brain[0],
        reference_label="Reference EEG",
        highlight_spans=spans,
        show=False,
    )

    plt.show()



.. image-sg:: /auto_examples/asr/images/sphx_glr_plot_02_mne_raw_qc_001.png
   :alt: ASR repairs plus optional final window rejection mask
   :srcset: /auto_examples/asr/images/sphx_glr_plot_02_mne_raw_qc_001.png
   :class: sphx-glr-single-img






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

   **Total running time of the script:** (0 minutes 0.918 seconds)


.. _sphx_glr_download_auto_examples_asr_plot_02_mne_raw_qc.py:

.. only:: html

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

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

      :download:`Download Jupyter notebook: plot_02_mne_raw_qc.ipynb <plot_02_mne_raw_qc.ipynb>`

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

      :download:`Download Python source code: plot_02_mne_raw_qc.py <plot_02_mne_raw_qc.py>`

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

      :download:`Download zipped: plot_02_mne_raw_qc.zip <plot_02_mne_raw_qc.zip>`