.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/asr/plot_11_epochs_and_meg.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_11_epochs_and_meg.py>`
        to download the full example code.

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

.. _sphx_glr_auto_examples_asr_plot_11_epochs_and_meg.py:


ASR on Epochs and on MEG.
=========================

The same :class:`~mne_denoise.asr.ASR` estimator works across MNE container
types and channel types. This example shows two cases the other examples do not:

1. **Epochs** --- calibrate on continuous ``Raw`` and clean segmented
   ``mne.Epochs`` with the fitted model.
2. **MEG** --- ASR is unit/scale agnostic, so ``picks="mag"`` cleans
   magnetometers exactly as ``picks="eeg"`` cleans EEG.

.. GENERATED FROM PYTHON SOURCE LINES 15-17

Part 1 - Epochs: fit on Raw, transform Epochs
---------------------------------------------

.. GENERATED FROM PYTHON SOURCE LINES 17-72

.. 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

    rng = np.random.default_rng(5)
    sfreq = 250.0
    n_channels, n_times = 8, 12000  # 48 s
    t = np.arange(n_times) / sfreq
    brain = np.vstack(
        [
            0.6 * np.sin(2 * np.pi * 10.0 * t + rng.uniform(0, 6.28))
            + 0.05 * rng.standard_normal(n_times)
            for _ in range(n_channels)
        ]
    )
    contaminated = brain.copy()
    for start in np.linspace(500, n_times - 500, 9).astype(int):
        spatial = rng.standard_normal(n_channels)
        spatial /= np.linalg.norm(spatial)
        contaminated[:, start : start + 200] += 10.0 * np.outer(
            spatial, rng.standard_normal(200)
        )

    info = mne.create_info([f"EEG{i:02d}" for i in range(n_channels)], sfreq, "eeg")
    raw = mne.io.RawArray(contaminated, info, verbose="ERROR")
    events = mne.make_fixed_length_events(raw, duration=2.0)
    epochs = mne.Epochs(
        raw, events, tmin=0.0, tmax=2.0, baseline=None, preload=True, verbose="ERROR"
    )

    # Calibrate on the continuous data, then clean the epochs.
    asr = ASR(cutoff=20.0, picks="eeg", verbose=False).fit(raw)
    epochs_clean = asr.transform(epochs)
    print(
        f"Epochs in: {epochs.get_data().shape} -> cleaned: {epochs_clean.get_data().shape}"
    )

    before = epochs.get_data()[:, 0].ravel()
    after = epochs_clean.get_data()[:, 0].ravel()
    plot_signal_overlay(
        before,
        after,
        np.arange(before.size) / sfreq,
        scale_after=False,
        before_label="epochs (raw)",
        after_label="epochs (ASR)",
        x_label="concatenated epoch time (s)",
        y_label="Amplitude (a.u.)",
        title="ASR on mne.Epochs (channel 0)",
        show=False,
    )




.. image-sg:: /auto_examples/asr/images/sphx_glr_plot_11_epochs_and_meg_001.png
   :alt: ASR on mne.Epochs (channel 0)
   :srcset: /auto_examples/asr/images/sphx_glr_plot_11_epochs_and_meg_001.png
   :class: sphx-glr-single-img


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

 .. code-block:: none

    Epochs in: (23, 8, 501) -> cleaned: (23, 8, 501)

    <Figure size 2400x800 with 1 Axes>



.. GENERATED FROM PYTHON SOURCE LINES 73-76

Part 2 - MEG: clean magnetometers
---------------------------------
ASR is scale-agnostic, so MEG (Tesla-scale) is handled like EEG.

.. GENERATED FROM PYTHON SOURCE LINES 76-106

.. code-block:: Python

    sample = mne.datasets.sample.data_path()
    raw_meg = mne.io.read_raw_fif(
        sample / "MEG" / "sample" / "sample_audvis_raw.fif", preload=True, verbose="ERROR"
    )
    raw_meg.pick("mag").crop(0, 40).resample(150, verbose="ERROR")
    raw_meg.filter(1.0, None, verbose="ERROR")

    asr_meg = ASR(cutoff=20.0, picks="mag", verbose=False)
    meg_clean = asr_meg.fit_transform(raw_meg.copy())

    var_before = float(np.var(raw_meg.get_data()))
    var_after = float(np.var(meg_clean.get_data()))
    print(f"MEG variance reduced: {100.0 * (1 - var_after / var_before):.1f}%")

    noisiest = int(np.argmax(np.var(raw_meg.get_data(), axis=1)))
    plot_signal_overlay(
        raw_meg,
        meg_clean,
        raw_meg.times,
        pick=raw_meg.ch_names[noisiest],
        scale_after=False,
        before_label="raw",
        after_label="ASR",
        x_label="Time (s)",
        y_label="Amplitude (T)",
        title=f"ASR on MEG magnetometer {raw_meg.ch_names[noisiest]}",
        show=False,
    )

    plt.show()



.. image-sg:: /auto_examples/asr/images/sphx_glr_plot_11_epochs_and_meg_002.png
   :alt: ASR on MEG magnetometer MEG 1411
   :srcset: /auto_examples/asr/images/sphx_glr_plot_11_epochs_and_meg_002.png
   :class: sphx-glr-single-img


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

 .. code-block:: none

    MEG variance reduced: 0.0%





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

   **Total running time of the script:** (0 minutes 7.841 seconds)


.. _sphx_glr_download_auto_examples_asr_plot_11_epochs_and_meg.py:

.. only:: html

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

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

      :download:`Download Jupyter notebook: plot_11_epochs_and_meg.ipynb <plot_11_epochs_and_meg.ipynb>`

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

      :download:`Download Python source code: plot_11_epochs_and_meg.py <plot_11_epochs_and_meg.py>`

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

      :download:`Download zipped: plot_11_epochs_and_meg.zip <plot_11_epochs_and_meg.zip>`