.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_workshop/02_eeg/plot_ERP.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

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

        Click :ref:`here <sphx_glr_download_auto_workshop_02_eeg_plot_ERP.py>`
        to download the full example code

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

.. _sphx_glr_auto_workshop_02_eeg_plot_ERP.py:


.. _compute_perp:

===============
02. Compute ERP
===============
This workflow mainly call the
:func:`ephypype pipeline <ephypype.pipelines.create_pipeline_evoked>`
computing N170 component from cleaned EEG data. The first Node of the workflow
(:ref:`extract_events_node` Node) extracts the events from raw data. The events
are saved in the Node directory.
In the :ref:`ERP_pipeline` the raw data are epoched accordingly to events
extracted in :ref:`extract_events` Node.
The evoked datasets are created by averaging the different conditions specified
in ``json`` file.

.. GENERATED FROM PYTHON SOURCE LINES 17-23

.. code-block:: default


    # Authors: Annalisa Pascarella <a.pascarella@iac.cnr.it>
    # License: BSD (3-clause)

    # sphinx_gallery_thumbnail_number = 2








.. GENERATED FROM PYTHON SOURCE LINES 24-36

Import modules
^^^^^^^^^^^^^^
The first step is to import the modules we need in the script. We import
mainly from |nipype| and |ephypype| packages.

.. |nipype| raw:: html

   <a href="https://nipype.readthedocs.io/en/latest/#" target="_blank">nipype</a>

.. |ephypype| raw:: html

   <a href="https://neuropycon.github.io/ephypype/index.html#ephypype target="_blank">ephypype</a>

.. GENERATED FROM PYTHON SOURCE LINES 36-51

.. code-block:: default


    import os
    import os.path as op

    import json
    import pprint  # noqa
    import ephypype

    import nipype.pipeline.engine as pe
    from nipype.interfaces.utility import Function

    from ephypype.nodes import create_iterator, create_datagrabber
    from ephypype.pipelines.preproc_meeg import create_pipeline_evoked
    from ephypype.datasets import fetch_erpcore_dataset








.. GENERATED FROM PYTHON SOURCE LINES 52-53

Let us fetch the data first. It is around 90 MB download.

.. GENERATED FROM PYTHON SOURCE LINES 53-66

.. code-block:: default

    import ephypype
    home_dir = op.expanduser("~")

    base_path = op.join(home_dir, 'workshop')

    try:
        os.mkdir(base_path)

    except OSError:
        print("directory {} already exists".format(base_path))

    data_path = fetch_erpcore_dataset(base_path)





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

 .. code-block:: none

    directory /home/pasca/workshop already exists




.. GENERATED FROM PYTHON SOURCE LINES 67-75

Define data and variables
^^^^^^^^^^^^^^^^^^^^^^^^^
Let us specify the variables that are specific for the data analysis (the
main directories where the data are stored, the list of subjects and
sessions, ...) and the variable specific for the particular pipeline
(events_id, baseline, ...) in a 
:download:`json <https://github.com/neuropycon/ephypype/tree/master/doc/workshop/eeg/params.json>` file 
(if it is does work, try to go on the github page, and right-click "Save As" on the Raw button)

.. GENERATED FROM PYTHON SOURCE LINES 75-97

.. code-block:: default


    # Read experiment params as json
    params = json.load(open("params.json"))
    pprint.pprint({'parameters': params["general"]})

    data_type = params["general"]["data_type"]
    subject_ids = params["general"]["subject_ids"]
    NJOBS = params["general"]["NJOBS"]
    session_ids = params["general"]["session_ids"]
    # data_path = params["general"]["data_path"]

    # ERP params
    ERP_str = 'ERP'
    pprint.pprint({'ERP': params[ERP_str]})
    events_id = params[ERP_str]['events_id']
    condition = params[ERP_str]['condition']
    baseline = tuple(params[ERP_str]['baseline'])
    events_file = params[ERP_str]['events_file']
    t_min = params[ERP_str]['tmin']
    t_max = params[ERP_str]['tmax']






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

 .. code-block:: none

    {'parameters': {'NJOBS': 1,
                    'data_path': '',
                    'data_type': 'eeg',
                    'session_ids': ['N170'],
                    'subject_ids': ['016']}}
    {'ERP': {'baseline': [None, 0],
             'condition': ['faces', 'car'],
             'events_file': '*_filt_ica-eve.fif',
             'events_id': {'car': 2, 'faces': 1},
             'tmax': 0.8,
             'tmin': -0.2}}




.. GENERATED FROM PYTHON SOURCE LINES 98-106

.. _extract_events:

Extract events
^^^^^^^^^^^^^^
The first Node of the workflow extract events from the raw data. The events
are extracted using the function `events_from_annotations <https://mne.tools/stable/generated/mne.events_from_annotations.html>`_
of `MNE-python <https://mne.tools/stable/index.html>`_ on raw data.
The events are saved in the Node directory.

.. GENERATED FROM PYTHON SOURCE LINES 106-155

.. code-block:: default

    def get_events(raw_ica, subject):
        '''
        First, we get the ica file from the preprocessing workflow directory, i.e.
        the cleaned raw data. The events are extracted from raw annotation and are
        saved in the Node directory.
        '''
        print(subject, raw_ica)
        import mne

        rename_events = {
            '201': 'response/correct',
            '202': 'response/error'
        }

        for i in range(1, 180 + 1):
            orig_name = f'{i}'

            if 1 <= i <= 40:
                new_name = 'stimulus/face/normal'
            elif 41 <= i <= 80:
                new_name = 'stimulus/car/normal'
            elif 101 <= i <= 140:
                new_name = 'stimulus/face/scrambled'
            elif 141 <= i <= 180:
                new_name = 'stimulus/car/scrambled'
            else:
                continue

            rename_events[orig_name] = new_name

        raw = mne.io.read_raw_fif(raw_ica, preload=True)
        events_from_annot, event_dict = mne.events_from_annotations(raw)

        faces = list()
        car = list()
        for key in event_dict.keys():
            if rename_events[key] == 'stimulus/car/normal':
                car.append(event_dict[key])
            elif rename_events[key] == 'stimulus/face/normal':
                faces.append(event_dict[key])

        merged_events = mne.merge_events(events_from_annot, faces, 1)
        merged_events = mne.merge_events(merged_events, car, 2)

        event_file = raw_ica.replace('.fif', '-eve.fif')
        mne.write_events(event_file, merged_events)

        return event_file








.. GENERATED FROM PYTHON SOURCE LINES 156-167

Specify Nodes
^^^^^^^^^^^^^

Before to create a workflow we have to create the `nodes <https://miykael.github.io/nipype_tutorial/notebooks/basic_nodes.html>`_
that define the workflow itself. In this example the main Nodes are

* ``infosource`` is a Node that just distributes values
* ``datasource`` is a DataGrabber Node that allows the user to define flexible search patterns which can be parameterized by user defined inputs 
* ``extract_events`` is a Node containing the function :ref:`extract_events`
* ``ERP_pipeline`` is a Node containing the pipeline created by :func:`~ephypype.pipelines.create_pipeline_evoked`


.. GENERATED FROM PYTHON SOURCE LINES 169-172

Infosource and Datasource
"""""""""""""""""""""""""
We create a node to pass input filenames to

.. GENERATED FROM PYTHON SOURCE LINES 172-175

.. code-block:: default

    infosource = create_iterator(['subject_id', 'session_id'],
                                 [subject_ids, session_ids])








.. GENERATED FROM PYTHON SOURCE LINES 176-185

.. code-block:: default


    # ``datasource`` node to grab data. The ``template_args`` in this node iterate
    # upon the values in the infosource node
    ica_dir = op.join(
        data_path, 'preprocessing_workflow', 'preproc_eeg_pipeline')
    template_path = "_session_id_%s_subject_id_%s/ica/sub-%s_ses-%s_*filt_ica.fif"
    template_args = [['session_id', 'subject_id', 'subject_id', 'session_id']]
    datasource = create_datagrabber(ica_dir, template_path, template_args)








.. GENERATED FROM PYTHON SOURCE LINES 186-191

.. _extract_events_node:

Extract events Node
"""""""""""""""""""
Then, we define the Node that encapsulates ``get_events`` function

.. GENERATED FROM PYTHON SOURCE LINES 191-197

.. code-block:: default

    extract_events = pe.Node(
        Function(input_names=['raw_ica', 'subject'],
                 output_names=['event_file'],
                 function=get_events),
        name='extract_events')








.. GENERATED FROM PYTHON SOURCE LINES 198-204

.. _ERP_pipeline:

ERP Node
""""""""
Finally, we create the ephypype pipeline computing evoked data which can be
connected to these nodes we created.

.. GENERATED FROM PYTHON SOURCE LINES 204-209

.. code-block:: default

    ERP_workflow = create_pipeline_evoked(
        data_path, data_type=data_type, pipeline_name="ERP_pipeline",
        events_id=events_id, baseline=baseline,
        condition=condition, t_min=t_min, t_max=t_max)








.. GENERATED FROM PYTHON SOURCE LINES 210-214

Specify Workflows and Connect Nodes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Now, we create our workflow and specify the ``base_dir`` which tells
nipype the directory in which to store the outputs.

.. GENERATED FROM PYTHON SOURCE LINES 214-220

.. code-block:: default


    ERP_pipeline_name = ERP_str + '_workflow'

    main_workflow = pe.Workflow(name=ERP_pipeline_name)
    main_workflow.base_dir = data_path








.. GENERATED FROM PYTHON SOURCE LINES 221-223

We then connect the output of ``infosource`` node to the one of
``datasource``. So, these two nodes taken together can grab data.

.. GENERATED FROM PYTHON SOURCE LINES 223-226

.. code-block:: default

    main_workflow.connect(infosource, 'subject_id', datasource,  'subject_id')
    main_workflow.connect(infosource, 'session_id', datasource, 'session_id')








.. GENERATED FROM PYTHON SOURCE LINES 227-229

We connect the output of ``infosource`` and ``datasource`` to the input
of ``extract_events`` node

.. GENERATED FROM PYTHON SOURCE LINES 229-232

.. code-block:: default

    main_workflow.connect(datasource, 'raw_file', extract_events, 'raw_ica')
    main_workflow.connect(infosource, 'subject_id', extract_events, 'subject')








.. GENERATED FROM PYTHON SOURCE LINES 233-235

Finally, we connect the output of ``infosource``, ``datasource`` and
``extract_events`` nodes to the input of ``ERP_pipeline`` node.

.. GENERATED FROM PYTHON SOURCE LINES 235-242

.. code-block:: default

    main_workflow.connect(infosource, 'subject_id',
                          ERP_workflow, 'inputnode.sbj_id')
    main_workflow.connect(datasource, 'raw_file',
                          ERP_workflow, 'inputnode.raw')
    main_workflow.connect(extract_events, 'event_file',
                          ERP_workflow, 'inputnode.events_file')








.. GENERATED FROM PYTHON SOURCE LINES 243-249

Run workflow
^^^^^^^^^^^^
After we have specified all the nodes and connections of the workflow, the
last step is to run it by calling the ``run`` method. It’s also possible to
generate static graph representing nodes and connections between them by
calling ``write_graph`` method.

.. GENERATED FROM PYTHON SOURCE LINES 249-251

.. code-block:: default

    main_workflow.write_graph(graph2use='colored')  # optional





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

 .. code-block:: none


    '/home/pasca/workshop/ERP_CORE/ERP_workflow/graph.png'



.. GENERATED FROM PYTHON SOURCE LINES 252-254

Take a moment to pause and notice how the connections
here correspond to how we connected the nodes.

.. GENERATED FROM PYTHON SOURCE LINES 254-261

.. code-block:: default


    import matplotlib.pyplot as plt  # noqa
    img = plt.imread(op.join(data_path, ERP_pipeline_name, 'graph.png'))
    plt.figure(figsize=(6, 6))
    plt.imshow(img)
    plt.axis('off')




.. image-sg:: /auto_workshop/02_eeg/images/sphx_glr_plot_ERP_001.png
   :alt: plot ERP
   :srcset: /auto_workshop/02_eeg/images/sphx_glr_plot_ERP_001.png
   :class: sphx-glr-single-img


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

 .. code-block:: none


    (-0.5, 535.5, 594.5, -0.5)



.. GENERATED FROM PYTHON SOURCE LINES 262-263

Finally, we are now ready to execute our workflow.

.. GENERATED FROM PYTHON SOURCE LINES 263-267

.. code-block:: default

    main_workflow.config['execution'] = {'remove_unnecessary_outputs': 'false'}
    # Run workflow locally on 1 CPU
    main_workflow.run(plugin='LegacyMultiProc', plugin_args={'n_procs': NJOBS})





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

 .. code-block:: none


    <networkx.classes.digraph.DiGraph object at 0x7fdef9f2e410>



.. GENERATED FROM PYTHON SOURCE LINES 268-270

Plot results
^^^^^^^^^^^^

.. GENERATED FROM PYTHON SOURCE LINES 270-297

.. code-block:: default

    import mne  # noqa
    import matplotlib.pyplot as plt  # noqa
    from ephypype.gather import get_results  # noqa

    evoked_files, _ = get_results(main_workflow.base_dir,
                                  main_workflow.name, pipeline='compute_evoked')

    for evoked_file in evoked_files:
        print(f'*** {evoked_file} ***\n')

        ave = mne.read_evokeds(evoked_file)
        faces, car = ave[0], ave[1]

        gfp_faces = faces.data.std(axis=0, ddof=0)
        gfp_car = car.data.std(axis=0, ddof=0)

        # compare conditions
        contrast = mne.combine_evoked([faces, car], weights=[1, -1])
        gfp_contrast = contrast.data.std(axis=0, ddof=0)

        # Reproducing the MNE-Python plot style seen above
        fig, ax = plt.subplots()
        ax.plot(faces.times, gfp_faces * 1e6, color='blue')
        ax.plot(car.times, gfp_car * 1e6, color='orange')
        ax.plot(contrast.times, gfp_contrast * 1e6, color='green')
        ax.legend(['Faces', 'Car', 'Contrast'])
        fig.show()



.. image-sg:: /auto_workshop/02_eeg/images/sphx_glr_plot_ERP_002.png
   :alt: plot ERP
   :srcset: /auto_workshop/02_eeg/images/sphx_glr_plot_ERP_002.png
   :class: sphx-glr-single-img


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

 .. code-block:: none

    *** /home/pasca/workshop/ERP_CORE/ERP_workflow/ERP_pipeline/_session_id_N170_subject_id_016/compute_evoked/sub-016_ses-N170_task-N170_eeg_filt_ica-ave.fif ***

    Reading /home/pasca/workshop/ERP_CORE/ERP_workflow/ERP_pipeline/_session_id_N170_subject_id_016/compute_evoked/sub-016_ses-N170_task-N170_eeg_filt_ica-ave.fif ...
        Read a total of 1 projection items:
            Average EEG reference (1 x 29) active
        Found the data of interest:
            t =    -200.20 ...     799.80 ms (faces)
            0 CTF compensation matrices available
            nave = 70 - aspect type = 100
    Projections have already been applied. Setting proj attribute to True.
    Loaded Evoked data is baseline-corrected (baseline: [-0.200195, 0] sec)
        Read a total of 1 projection items:
            Average EEG reference (1 x 29) active
        Found the data of interest:
            t =    -200.20 ...     799.80 ms (car)
            0 CTF compensation matrices available
            nave = 76 - aspect type = 100
    Projections have already been applied. Setting proj attribute to True.
    Loaded Evoked data is baseline-corrected (baseline: [-0.200195, 0] sec)





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

   **Total running time of the script:** ( 0 minutes  8.560 seconds)


.. _sphx_glr_download_auto_workshop_02_eeg_plot_ERP.py:

.. only:: html

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


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

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

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

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


.. only:: html

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

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