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)
[Parallel(n_jobs=1)]: Done 17 tasks | elapsed: 0.0s
[Parallel(n_jobs=1)]: Done 71 tasks | elapsed: 0.1s
[Parallel(n_jobs=1)]: Done 161 tasks | elapsed: 0.3s
[Parallel(n_jobs=1)]: Done 287 tasks | elapsed: 0.6s
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, 54.17it/s]
11%|█▏ | Fitting GeneralizingEstimator : 4/35 [00:00<00:00, 56.70it/s]
17%|█▋ | Fitting GeneralizingEstimator : 6/35 [00:00<00:00, 57.60it/s]
26%|██▌ | Fitting GeneralizingEstimator : 9/35 [00:00<00:00, 65.64it/s]
34%|███▍ | Fitting GeneralizingEstimator : 12/35 [00:00<00:00, 70.70it/s]
43%|████▎ | Fitting GeneralizingEstimator : 15/35 [00:00<00:00, 74.11it/s]
51%|█████▏ | Fitting GeneralizingEstimator : 18/35 [00:00<00:00, 76.55it/s]
60%|██████ | Fitting GeneralizingEstimator : 21/35 [00:00<00:00, 78.38it/s]
66%|██████▌ | Fitting GeneralizingEstimator : 23/35 [00:00<00:00, 75.83it/s]
74%|███████▍ | Fitting GeneralizingEstimator : 26/35 [00:00<00:00, 77.46it/s]
80%|████████ | Fitting GeneralizingEstimator : 28/35 [00:00<00:00, 74.74it/s]
89%|████████▊ | Fitting GeneralizingEstimator : 31/35 [00:00<00:00, 76.27it/s]
97%|█████████▋| Fitting GeneralizingEstimator : 34/35 [00:00<00:00, 77.56it/s]
100%|██████████| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 76.94it/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 : 13/1225 [00:00<00:03, 354.57it/s]
2%|▏ | Scoring GeneralizingEstimator : 25/1225 [00:00<00:03, 355.10it/s]
3%|▎ | Scoring GeneralizingEstimator : 37/1225 [00:00<00:03, 355.06it/s]
4%|▍ | Scoring GeneralizingEstimator : 50/1225 [00:00<00:03, 363.01it/s]
5%|▌ | Scoring GeneralizingEstimator : 62/1225 [00:00<00:03, 361.55it/s]
6%|▌ | Scoring GeneralizingEstimator : 74/1225 [00:00<00:03, 360.33it/s]
7%|▋ | Scoring GeneralizingEstimator : 87/1225 [00:00<00:03, 364.45it/s]
8%|▊ | Scoring GeneralizingEstimator : 99/1225 [00:00<00:03, 362.89it/s]
9%|▉ | Scoring GeneralizingEstimator : 112/1225 [00:00<00:03, 365.79it/s]
10%|█ | Scoring GeneralizingEstimator : 125/1225 [00:00<00:02, 367.96it/s]
11%|█ | Scoring GeneralizingEstimator : 137/1225 [00:00<00:02, 366.61it/s]
12%|█▏ | Scoring GeneralizingEstimator : 150/1225 [00:00<00:02, 368.15it/s]
13%|█▎ | Scoring GeneralizingEstimator : 163/1225 [00:00<00:02, 367.41it/s]
14%|█▍ | Scoring GeneralizingEstimator : 176/1225 [00:00<00:02, 369.03it/s]
15%|█▌ | Scoring GeneralizingEstimator : 188/1225 [00:00<00:02, 367.81it/s]
16%|█▋ | Scoring GeneralizingEstimator : 201/1225 [00:00<00:02, 369.37it/s]
17%|█▋ | Scoring GeneralizingEstimator : 213/1225 [00:00<00:02, 368.27it/s]
18%|█▊ | Scoring GeneralizingEstimator : 226/1225 [00:00<00:02, 369.64it/s]
19%|█▉ | Scoring GeneralizingEstimator : 238/1225 [00:00<00:02, 368.45it/s]
20%|██ | Scoring GeneralizingEstimator : 251/1225 [00:00<00:02, 369.72it/s]
21%|██▏ | Scoring GeneralizingEstimator : 263/1225 [00:00<00:02, 368.62it/s]
23%|██▎ | Scoring GeneralizingEstimator : 277/1225 [00:00<00:02, 370.89it/s]
24%|██▎ | Scoring GeneralizingEstimator : 289/1225 [00:00<00:02, 369.79it/s]
25%|██▍ | Scoring GeneralizingEstimator : 302/1225 [00:00<00:02, 370.93it/s]
26%|██▌ | Scoring GeneralizingEstimator : 314/1225 [00:00<00:02, 369.89it/s]
27%|██▋ | Scoring GeneralizingEstimator : 327/1225 [00:00<00:02, 369.59it/s]
28%|██▊ | Scoring GeneralizingEstimator : 340/1225 [00:00<00:02, 370.69it/s]
29%|██▊ | Scoring GeneralizingEstimator : 352/1225 [00:00<00:02, 369.79it/s]
30%|██▉ | Scoring GeneralizingEstimator : 365/1225 [00:00<00:02, 370.77it/s]
31%|███ | Scoring GeneralizingEstimator : 378/1225 [00:01<00:02, 371.58it/s]
32%|███▏ | Scoring GeneralizingEstimator : 390/1225 [00:01<00:02, 370.59it/s]
33%|███▎ | Scoring GeneralizingEstimator : 403/1225 [00:01<00:02, 370.40it/s]
34%|███▍ | Scoring GeneralizingEstimator : 416/1225 [00:01<00:02, 371.34it/s]
35%|███▍ | Scoring GeneralizingEstimator : 428/1225 [00:01<00:02, 370.36it/s]
36%|███▌ | Scoring GeneralizingEstimator : 441/1225 [00:01<00:02, 371.16it/s]
37%|███▋ | Scoring GeneralizingEstimator : 453/1225 [00:01<00:02, 370.25it/s]
38%|███▊ | Scoring GeneralizingEstimator : 466/1225 [00:01<00:02, 371.00it/s]
39%|███▉ | Scoring GeneralizingEstimator : 479/1225 [00:01<00:02, 369.76it/s]
40%|████ | Scoring GeneralizingEstimator : 492/1225 [00:01<00:01, 370.69it/s]
41%|████ | Scoring GeneralizingEstimator : 505/1225 [00:01<00:01, 371.31it/s]
42%|████▏ | Scoring GeneralizingEstimator : 517/1225 [00:01<00:01, 370.38it/s]
43%|████▎ | Scoring GeneralizingEstimator : 530/1225 [00:01<00:01, 371.24it/s]
44%|████▍ | Scoring GeneralizingEstimator : 542/1225 [00:01<00:01, 370.35it/s]
45%|████▌ | Scoring GeneralizingEstimator : 555/1225 [00:01<00:01, 371.23it/s]
46%|████▋ | Scoring GeneralizingEstimator : 567/1225 [00:01<00:01, 370.34it/s]
47%|████▋ | Scoring GeneralizingEstimator : 580/1225 [00:01<00:01, 371.17it/s]
48%|████▊ | Scoring GeneralizingEstimator : 592/1225 [00:01<00:01, 370.29it/s]
49%|████▉ | Scoring GeneralizingEstimator : 605/1225 [00:01<00:01, 371.05it/s]
50%|█████ | Scoring GeneralizingEstimator : 618/1225 [00:01<00:01, 370.30it/s]
52%|█████▏ | Scoring GeneralizingEstimator : 631/1225 [00:01<00:01, 371.01it/s]
52%|█████▏ | Scoring GeneralizingEstimator : 643/1225 [00:01<00:01, 370.04it/s]
54%|█████▎ | Scoring GeneralizingEstimator : 656/1225 [00:01<00:01, 370.71it/s]
55%|█████▍ | Scoring GeneralizingEstimator : 668/1225 [00:01<00:01, 369.86it/s]
56%|█████▌ | Scoring GeneralizingEstimator : 681/1225 [00:01<00:01, 370.72it/s]
57%|█████▋ | Scoring GeneralizingEstimator : 693/1225 [00:01<00:01, 369.94it/s]
58%|█████▊ | Scoring GeneralizingEstimator : 706/1225 [00:01<00:01, 370.81it/s]
59%|█████▊ | Scoring GeneralizingEstimator : 719/1225 [00:01<00:01, 370.77it/s]
60%|█████▉ | Scoring GeneralizingEstimator : 731/1225 [00:01<00:01, 369.99it/s]
61%|██████ | Scoring GeneralizingEstimator : 744/1225 [00:02<00:01, 370.82it/s]
62%|██████▏ | Scoring GeneralizingEstimator : 756/1225 [00:02<00:01, 369.90it/s]
63%|██████▎ | Scoring GeneralizingEstimator : 769/1225 [00:02<00:01, 370.74it/s]
64%|██████▍ | Scoring GeneralizingEstimator : 782/1225 [00:02<00:01, 370.28it/s]
65%|██████▍ | Scoring GeneralizingEstimator : 795/1225 [00:02<00:01, 370.99it/s]
66%|██████▌ | Scoring GeneralizingEstimator : 807/1225 [00:02<00:01, 370.23it/s]
67%|██████▋ | Scoring GeneralizingEstimator : 820/1225 [00:02<00:01, 371.00it/s]
68%|██████▊ | Scoring GeneralizingEstimator : 832/1225 [00:02<00:01, 370.24it/s]
69%|██████▉ | Scoring GeneralizingEstimator : 845/1225 [00:02<00:01, 370.99it/s]
70%|██████▉ | Scoring GeneralizingEstimator : 857/1225 [00:02<00:00, 370.20it/s]
71%|███████ | Scoring GeneralizingEstimator : 870/1225 [00:02<00:00, 371.00it/s]
72%|███████▏ | Scoring GeneralizingEstimator : 883/1225 [00:02<00:00, 370.60it/s]
73%|███████▎ | Scoring GeneralizingEstimator : 896/1225 [00:02<00:00, 371.37it/s]
74%|███████▍ | Scoring GeneralizingEstimator : 908/1225 [00:02<00:00, 370.57it/s]
75%|███████▌ | Scoring GeneralizingEstimator : 920/1225 [00:02<00:00, 369.83it/s]
76%|███████▌ | Scoring GeneralizingEstimator : 931/1225 [00:02<00:00, 367.46it/s]
77%|███████▋ | Scoring GeneralizingEstimator : 943/1225 [00:02<00:00, 366.88it/s]
78%|███████▊ | Scoring GeneralizingEstimator : 956/1225 [00:02<00:00, 367.08it/s]
79%|███████▉ | Scoring GeneralizingEstimator : 969/1225 [00:02<00:00, 368.02it/s]
80%|████████ | Scoring GeneralizingEstimator : 982/1225 [00:02<00:00, 368.88it/s]
81%|████████ | Scoring GeneralizingEstimator : 994/1225 [00:02<00:00, 368.23it/s]
82%|████████▏ | Scoring GeneralizingEstimator : 1007/1225 [00:02<00:00, 369.04it/s]
83%|████████▎ | Scoring GeneralizingEstimator : 1021/1225 [00:02<00:00, 370.96it/s]
85%|████████▍ | Scoring GeneralizingEstimator : 1037/1225 [00:02<00:00, 375.09it/s]
86%|████████▌ | Scoring GeneralizingEstimator : 1052/1225 [00:02<00:00, 378.61it/s]
87%|████████▋ | Scoring GeneralizingEstimator : 1067/1225 [00:02<00:00, 381.93it/s]
88%|████████▊ | Scoring GeneralizingEstimator : 1082/1225 [00:02<00:00, 385.07it/s]
90%|████████▉ | Scoring GeneralizingEstimator : 1097/1225 [00:02<00:00, 388.11it/s]
91%|█████████ | Scoring GeneralizingEstimator : 1112/1225 [00:02<00:00, 390.94it/s]
92%|█████████▏| Scoring GeneralizingEstimator : 1127/1225 [00:02<00:00, 393.66it/s]
93%|█████████▎| Scoring GeneralizingEstimator : 1143/1225 [00:03<00:00, 396.62it/s]
95%|█████████▍| Scoring GeneralizingEstimator : 1159/1225 [00:03<00:00, 400.54it/s]
96%|█████████▌| Scoring GeneralizingEstimator : 1174/1225 [00:03<00:00, 402.75it/s]
97%|█████████▋| Scoring GeneralizingEstimator : 1190/1225 [00:03<00:00, 406.22it/s]
98%|█████████▊| Scoring GeneralizingEstimator : 1205/1225 [00:03<00:00, 408.14it/s]
100%|█████████▉| Scoring GeneralizingEstimator : 1220/1225 [00:03<00:00, 409.98it/s]
100%|██████████| Scoring GeneralizingEstimator : 1225/1225 [00:03<00:00, 381.45it/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 5.612 seconds)