Convert EEG MNE_RawArray to ndarray with time courses

✉️ Mailing List Archive (read-only)
1

External Email - Use Caution

Hi,
Basically I have an MNE RawArray object and I would like to convert it to
an ndarray.

I am doing an ICA of an EEG signal with 22 channels. I need an ndarray with
the corresponding time courses of the the Independent Components from the
ICA.

Code Snippet
ica = mne.preprocessing.ICA(method="infomax", random_state=1)
ica.fit(raw_tmp)

data = ica.get_sources(inst=raw_tmp)

data is a RawArray object
Using
cpt = data.get_data()

gives an ndarray but without the time courses of the ICs

Thank you for any help.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20190827/d72d69ac/attachment.html

2

External Email - Use Caution

This works for me:

import os
import numpy as np
import matplotlib.pyplot as plt
import mne
from mne.preprocessing import ICA

sample_data_folder = mne.datasets.sample.data_path()
sample_data_raw_file = os.path.join(sample_data_folder, 'MEG', 'sample',
'sample_audvis_raw.fif')
raw = mne.io.read_raw_fif(sample_data_raw_file)
raw.crop(tmax=60.).load_data().filter(l_freq=1., h_freq=None)
ica = ICA(n_components=30, random_state=97)
ica.fit(raw)

sources_raw = ica.get_sources(raw)
sources_array, times = sources_raw.get_data(return_times=True)

# PLOT THE NUMPY ARRAY OF SOURCES
fig, ax = plt.subplots()
# this will let us plot the sources without overlapping:
offsets = 15 * np.linspace(29, 0, len(sources_array))
ax.plot(times, sources_array.T + offsets, linewidth=0.5, color='k')

# COMPARE WITH THE MNE-PYTHON PLOT OF SOURCES
ica.plot_sources(raw)