Note
Go to the end to download the full example code.
Decoding sensor space data with generalization across time and conditions#
This example runs the analysis described in [1]. It illustrates how one can fit a linear classifier to identify a discriminatory topography at a given time instant and subsequently assess whether this linear model can accurately predict all of the time samples of a second set of conditions.
# Authors: Jean-Rémi King <jeanremi.king@gmail.com>
# Alexandre Gramfort <alexandre.gramfort@inria.fr>
# Denis Engemann <denis.engemann@gmail.com>
#
# License: BSD-3-Clause
# Copyright the MNE-Python contributors.
import matplotlib.pyplot as plt
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler
import mne
from mne.datasets import sample
from mne.decoding import GeneralizingEstimator
print(__doc__)
# Preprocess data
data_path = sample.data_path()
# Load and filter data, set up epochs
meg_path = data_path / "MEG" / "sample"
raw_fname = meg_path / "sample_audvis_filt-0-40_raw.fif"
events_fname = meg_path / "sample_audvis_filt-0-40_raw-eve.fif"
raw = mne.io.read_raw_fif(raw_fname, preload=True)
picks = mne.pick_types(raw.info, meg=True, exclude="bads") # Pick MEG channels
raw.filter(1.0, 30.0, fir_design="firwin") # Band pass filtering signals
events = mne.read_events(events_fname)
event_id = {
"Auditory/Left": 1,
"Auditory/Right": 2,
"Visual/Left": 3,
"Visual/Right": 4,
}
tmin = -0.050
tmax = 0.400
# decimate to make the example faster to run, but then use verbose='error' in
# the Epochs constructor to suppress warning about decimation causing aliasing
decim = 2
epochs = mne.Epochs(
raw,
events,
event_id=event_id,
tmin=tmin,
tmax=tmax,
proj=True,
picks=picks,
baseline=None,
preload=True,
reject=dict(mag=5e-12),
decim=decim,
verbose="error",
)
Opening raw data file /home/circleci/mne_data/MNE-sample-data/MEG/sample/sample_audvis_filt-0-40_raw.fif...
Read a total of 4 projection items:
PCA-v1 (1 x 102) idle
PCA-v2 (1 x 102) idle
PCA-v3 (1 x 102) idle
Average EEG reference (1 x 60) idle
Range : 6450 ... 48149 = 42.956 ... 320.665 secs
Ready.
Reading 0 ... 41699 = 0.000 ... 277.709 secs...
Filtering raw data in 1 contiguous segment
Setting up band-pass filter from 1 - 30 Hz
FIR filter parameters
---------------------
Designing a one-pass, zero-phase, non-causal bandpass filter:
- Windowed time-domain design (firwin) method
- Hamming window with 0.0194 passband ripple and 53 dB stopband attenuation
- Lower passband edge: 1.00
- Lower transition bandwidth: 1.00 Hz (-6 dB cutoff frequency: 0.50 Hz)
- Upper passband edge: 30.00 Hz
- Upper transition bandwidth: 7.50 Hz (-6 dB cutoff frequency: 33.75 Hz)
- Filter length: 497 samples (3.310 s)
We will train the classifier on all left visual vs auditory trials and test on all right visual vs auditory trials.
clf = make_pipeline(
StandardScaler(),
LogisticRegression(solver="liblinear"), # liblinear is faster than lbfgs
)
time_gen = GeneralizingEstimator(clf, scoring="roc_auc", n_jobs=None, verbose=True)
# Fit classifiers on the epochs where the stimulus was presented to the left.
# Note that the experimental condition y indicates auditory or visual
time_gen.fit(X=epochs["Left"].get_data(copy=False), y=epochs["Left"].events[:, 2] > 2)
0%| | Fitting GeneralizingEstimator : 0/35 [00:00<?, ?it/s]
6%|▌ | Fitting GeneralizingEstimator : 2/35 [00:00<00:00, 54.71it/s]
11%|█▏ | Fitting GeneralizingEstimator : 4/35 [00:00<00:00, 57.03it/s]
20%|██ | Fitting GeneralizingEstimator : 7/35 [00:00<00:00, 67.99it/s]
26%|██▌ | Fitting GeneralizingEstimator : 9/35 [00:00<00:00, 65.73it/s]
37%|███▋ | Fitting GeneralizingEstimator : 13/35 [00:00<00:00, 77.31it/s]
46%|████▌ | Fitting GeneralizingEstimator : 16/35 [00:00<00:00, 79.52it/s]
54%|█████▍ | Fitting GeneralizingEstimator : 19/35 [00:00<00:00, 81.11it/s]
60%|██████ | Fitting GeneralizingEstimator : 21/35 [00:00<00:00, 77.92it/s]
69%|██████▊ | Fitting GeneralizingEstimator : 24/35 [00:00<00:00, 79.43it/s]
77%|███████▋ | Fitting GeneralizingEstimator : 27/35 [00:00<00:00, 80.64it/s]
83%|████████▎ | Fitting GeneralizingEstimator : 29/35 [00:00<00:00, 78.20it/s]
91%|█████████▏| Fitting GeneralizingEstimator : 32/35 [00:00<00:00, 78.89it/s]
100%|██████████| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 80.50it/s]
100%|██████████| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 79.52it/s]
Score on the epochs where the stimulus was presented to the right.
scores = time_gen.score(
X=epochs["Right"].get_data(copy=False), y=epochs["Right"].events[:, 2] > 2
)
0%| | Scoring GeneralizingEstimator : 0/1225 [00:00<?, ?it/s]
1%| | Scoring GeneralizingEstimator : 12/1225 [00:00<00:03, 331.38it/s]
2%|▏ | Scoring GeneralizingEstimator : 25/1225 [00:00<00:03, 358.67it/s]
3%|▎ | Scoring GeneralizingEstimator : 37/1225 [00:00<00:03, 357.85it/s]
4%|▍ | Scoring GeneralizingEstimator : 49/1225 [00:00<00:03, 357.33it/s]
5%|▍ | Scoring GeneralizingEstimator : 61/1225 [00:00<00:03, 357.10it/s]
6%|▌ | Scoring GeneralizingEstimator : 74/1225 [00:00<00:03, 362.35it/s]
7%|▋ | Scoring GeneralizingEstimator : 86/1225 [00:00<00:03, 360.94it/s]
8%|▊ | Scoring GeneralizingEstimator : 98/1225 [00:00<00:03, 360.32it/s]
9%|▉ | Scoring GeneralizingEstimator : 111/1225 [00:00<00:03, 363.54it/s]
10%|█ | Scoring GeneralizingEstimator : 123/1225 [00:00<00:03, 362.22it/s]
11%|█ | Scoring GeneralizingEstimator : 136/1225 [00:00<00:03, 362.54it/s]
12%|█▏ | Scoring GeneralizingEstimator : 148/1225 [00:00<00:02, 361.77it/s]
13%|█▎ | Scoring GeneralizingEstimator : 160/1225 [00:00<00:02, 361.10it/s]
14%|█▍ | Scoring GeneralizingEstimator : 173/1225 [00:00<00:02, 363.55it/s]
15%|█▌ | Scoring GeneralizingEstimator : 185/1225 [00:00<00:02, 362.87it/s]
16%|█▌ | Scoring GeneralizingEstimator : 198/1225 [00:00<00:02, 363.07it/s]
17%|█▋ | Scoring GeneralizingEstimator : 210/1225 [00:00<00:02, 362.46it/s]
18%|█▊ | Scoring GeneralizingEstimator : 223/1225 [00:00<00:02, 364.24it/s]
19%|█▉ | Scoring GeneralizingEstimator : 235/1225 [00:00<00:02, 363.46it/s]
20%|██ | Scoring GeneralizingEstimator : 247/1225 [00:00<00:02, 362.88it/s]
21%|██ | Scoring GeneralizingEstimator : 260/1225 [00:00<00:02, 364.61it/s]
22%|██▏ | Scoring GeneralizingEstimator : 272/1225 [00:00<00:02, 363.90it/s]
23%|██▎ | Scoring GeneralizingEstimator : 284/1225 [00:00<00:02, 363.25it/s]
24%|██▍ | Scoring GeneralizingEstimator : 296/1225 [00:00<00:02, 362.11it/s]
25%|██▌ | Scoring GeneralizingEstimator : 309/1225 [00:00<00:02, 363.76it/s]
26%|██▌ | Scoring GeneralizingEstimator : 321/1225 [00:00<00:02, 363.08it/s]
27%|██▋ | Scoring GeneralizingEstimator : 335/1225 [00:00<00:02, 364.22it/s]
28%|██▊ | Scoring GeneralizingEstimator : 347/1225 [00:00<00:02, 363.75it/s]
29%|██▉ | Scoring GeneralizingEstimator : 359/1225 [00:00<00:02, 363.30it/s]
30%|███ | Scoring GeneralizingEstimator : 371/1225 [00:01<00:02, 362.87it/s]
31%|███▏ | Scoring GeneralizingEstimator : 384/1225 [00:01<00:02, 364.27it/s]
32%|███▏ | Scoring GeneralizingEstimator : 396/1225 [00:01<00:02, 363.82it/s]
33%|███▎ | Scoring GeneralizingEstimator : 408/1225 [00:01<00:02, 363.36it/s]
34%|███▍ | Scoring GeneralizingEstimator : 420/1225 [00:01<00:02, 362.94it/s]
35%|███▌ | Scoring GeneralizingEstimator : 432/1225 [00:01<00:02, 362.54it/s]
36%|███▋ | Scoring GeneralizingEstimator : 445/1225 [00:01<00:02, 363.66it/s]
37%|███▋ | Scoring GeneralizingEstimator : 457/1225 [00:01<00:02, 363.20it/s]
38%|███▊ | Scoring GeneralizingEstimator : 469/1225 [00:01<00:02, 362.52it/s]
39%|███▉ | Scoring GeneralizingEstimator : 482/1225 [00:01<00:02, 362.29it/s]
40%|████ | Scoring GeneralizingEstimator : 495/1225 [00:01<00:02, 363.40it/s]
41%|████▏ | Scoring GeneralizingEstimator : 507/1225 [00:01<00:01, 362.89it/s]
42%|████▏ | Scoring GeneralizingEstimator : 519/1225 [00:01<00:01, 362.34it/s]
43%|████▎ | Scoring GeneralizingEstimator : 531/1225 [00:01<00:01, 361.98it/s]
44%|████▍ | Scoring GeneralizingEstimator : 543/1225 [00:01<00:01, 361.68it/s]
45%|████▌ | Scoring GeneralizingEstimator : 556/1225 [00:01<00:01, 363.05it/s]
46%|████▋ | Scoring GeneralizingEstimator : 568/1225 [00:01<00:01, 362.51it/s]
47%|████▋ | Scoring GeneralizingEstimator : 580/1225 [00:01<00:01, 362.10it/s]
48%|████▊ | Scoring GeneralizingEstimator : 593/1225 [00:01<00:01, 362.41it/s]
49%|████▉ | Scoring GeneralizingEstimator : 605/1225 [00:01<00:01, 361.88it/s]
50%|█████ | Scoring GeneralizingEstimator : 617/1225 [00:01<00:01, 361.60it/s]
51%|█████▏ | Scoring GeneralizingEstimator : 630/1225 [00:01<00:01, 362.73it/s]
52%|█████▏ | Scoring GeneralizingEstimator : 642/1225 [00:01<00:01, 362.12it/s]
53%|█████▎ | Scoring GeneralizingEstimator : 655/1225 [00:01<00:01, 362.54it/s]
54%|█████▍ | Scoring GeneralizingEstimator : 667/1225 [00:01<00:01, 362.15it/s]
55%|█████▌ | Scoring GeneralizingEstimator : 679/1225 [00:01<00:01, 361.78it/s]
56%|█████▋ | Scoring GeneralizingEstimator : 691/1225 [00:01<00:01, 361.50it/s]
57%|█████▋ | Scoring GeneralizingEstimator : 703/1225 [00:01<00:01, 361.19it/s]
59%|█████▊ | Scoring GeneralizingEstimator : 717/1225 [00:01<00:01, 362.24it/s]
60%|█████▉ | Scoring GeneralizingEstimator : 729/1225 [00:02<00:01, 361.91it/s]
60%|██████ | Scoring GeneralizingEstimator : 741/1225 [00:02<00:01, 361.56it/s]
61%|██████▏ | Scoring GeneralizingEstimator : 753/1225 [00:02<00:01, 361.26it/s]
62%|██████▏ | Scoring GeneralizingEstimator : 765/1225 [00:02<00:01, 361.01it/s]
64%|██████▎ | Scoring GeneralizingEstimator : 778/1225 [00:02<00:01, 362.30it/s]
64%|██████▍ | Scoring GeneralizingEstimator : 790/1225 [00:02<00:01, 361.85it/s]
65%|██████▌ | Scoring GeneralizingEstimator : 802/1225 [00:02<00:01, 361.59it/s]
66%|██████▋ | Scoring GeneralizingEstimator : 814/1225 [00:02<00:01, 361.30it/s]
67%|██████▋ | Scoring GeneralizingEstimator : 826/1225 [00:02<00:01, 361.03it/s]
68%|██████▊ | Scoring GeneralizingEstimator : 839/1225 [00:02<00:01, 362.18it/s]
69%|██████▉ | Scoring GeneralizingEstimator : 851/1225 [00:02<00:01, 361.85it/s]
71%|███████ | Scoring GeneralizingEstimator : 864/1225 [00:02<00:00, 361.33it/s]
72%|███████▏ | Scoring GeneralizingEstimator : 876/1225 [00:02<00:00, 361.09it/s]
72%|███████▏ | Scoring GeneralizingEstimator : 888/1225 [00:02<00:00, 360.77it/s]
73%|███████▎ | Scoring GeneralizingEstimator : 900/1225 [00:02<00:00, 360.49it/s]
74%|███████▍ | Scoring GeneralizingEstimator : 912/1225 [00:02<00:00, 360.25it/s]
76%|███████▌ | Scoring GeneralizingEstimator : 925/1225 [00:02<00:00, 361.58it/s]
76%|███████▋ | Scoring GeneralizingEstimator : 937/1225 [00:02<00:00, 361.31it/s]
77%|███████▋ | Scoring GeneralizingEstimator : 949/1225 [00:02<00:00, 361.05it/s]
78%|███████▊ | Scoring GeneralizingEstimator : 961/1225 [00:02<00:00, 360.83it/s]
79%|███████▉ | Scoring GeneralizingEstimator : 973/1225 [00:02<00:00, 360.55it/s]
80%|████████ | Scoring GeneralizingEstimator : 985/1225 [00:02<00:00, 360.32it/s]
81%|████████▏ | Scoring GeneralizingEstimator : 998/1225 [00:02<00:00, 361.53it/s]
82%|████████▏ | Scoring GeneralizingEstimator : 1010/1225 [00:02<00:00, 361.29it/s]
84%|████████▎ | Scoring GeneralizingEstimator : 1023/1225 [00:02<00:00, 361.12it/s]
85%|████████▍ | Scoring GeneralizingEstimator : 1036/1225 [00:02<00:00, 362.18it/s]
86%|████████▌ | Scoring GeneralizingEstimator : 1048/1225 [00:02<00:00, 361.85it/s]
87%|████████▋ | Scoring GeneralizingEstimator : 1060/1225 [00:02<00:00, 361.35it/s]
88%|████████▊ | Scoring GeneralizingEstimator : 1072/1225 [00:02<00:00, 360.95it/s]
88%|████████▊ | Scoring GeneralizingEstimator : 1084/1225 [00:02<00:00, 360.72it/s]
89%|████████▉ | Scoring GeneralizingEstimator : 1096/1225 [00:03<00:00, 360.49it/s]
91%|█████████ | Scoring GeneralizingEstimator : 1109/1225 [00:03<00:00, 361.74it/s]
92%|█████████▏| Scoring GeneralizingEstimator : 1121/1225 [00:03<00:00, 361.49it/s]
92%|█████████▏| Scoring GeneralizingEstimator : 1133/1225 [00:03<00:00, 360.31it/s]
94%|█████████▎| Scoring GeneralizingEstimator : 1146/1225 [00:03<00:00, 361.48it/s]
95%|█████████▍| Scoring GeneralizingEstimator : 1158/1225 [00:03<00:00, 361.23it/s]
96%|█████████▌| Scoring GeneralizingEstimator : 1170/1225 [00:03<00:00, 360.98it/s]
96%|█████████▋| Scoring GeneralizingEstimator : 1182/1225 [00:03<00:00, 360.75it/s]
97%|█████████▋| Scoring GeneralizingEstimator : 1194/1225 [00:03<00:00, 360.52it/s]
99%|█████████▊| Scoring GeneralizingEstimator : 1207/1225 [00:03<00:00, 361.81it/s]
100%|█████████▉| Scoring GeneralizingEstimator : 1219/1225 [00:03<00:00, 360.95it/s]
100%|██████████| Scoring GeneralizingEstimator : 1225/1225 [00:03<00:00, 361.79it/s]
Plot
fig, ax = plt.subplots(layout="constrained")
im = ax.matshow(
scores,
vmin=0,
vmax=1.0,
cmap="RdBu_r",
origin="lower",
extent=epochs.times[[0, -1, 0, -1]],
)
ax.axhline(0.0, color="k")
ax.axvline(0.0, color="k")
ax.xaxis.set_ticks_position("bottom")
ax.set_xlabel(
'Condition: "Right"\nTesting Time (s)',
)
ax.set_ylabel('Condition: "Left"\nTraining Time (s)')
ax.set_title("Generalization across time and condition", fontweight="bold")
fig.colorbar(im, ax=ax, label="Performance (ROC AUC)")
plt.show()

References#
Total running time of the script: (0 minutes 5.670 seconds)