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-Remi 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: 0.5s
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]
3%|▎ | Fitting GeneralizingEstimator : 1/35 [00:00<00:01, 29.23it/s]
11%|█▏ | Fitting GeneralizingEstimator : 4/35 [00:00<00:00, 59.25it/s]
17%|█▋ | Fitting GeneralizingEstimator : 6/35 [00:00<00:00, 59.28it/s]
26%|██▌ | Fitting GeneralizingEstimator : 9/35 [00:00<00:00, 67.09it/s]
34%|███▍ | Fitting GeneralizingEstimator : 12/35 [00:00<00:00, 71.89it/s]
46%|████▌ | Fitting GeneralizingEstimator : 16/35 [00:00<00:00, 80.44it/s]
54%|█████▍ | Fitting GeneralizingEstimator : 19/35 [00:00<00:00, 81.83it/s]
63%|██████▎ | Fitting GeneralizingEstimator : 22/35 [00:00<00:00, 82.89it/s]
71%|███████▏ | Fitting GeneralizingEstimator : 25/35 [00:00<00:00, 83.71it/s]
80%|████████ | Fitting GeneralizingEstimator : 28/35 [00:00<00:00, 84.27it/s]
86%|████████▌ | Fitting GeneralizingEstimator : 30/35 [00:00<00:00, 81.39it/s]
94%|█████████▍| Fitting GeneralizingEstimator : 33/35 [00:00<00:00, 82.22it/s]
100%|██████████| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 83.01it/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 : 14/1225 [00:00<00:02, 404.83it/s]
3%|▎ | Scoring GeneralizingEstimator : 31/1225 [00:00<00:02, 453.10it/s]
4%|▍ | Scoring GeneralizingEstimator : 48/1225 [00:00<00:02, 470.15it/s]
5%|▌ | Scoring GeneralizingEstimator : 63/1225 [00:00<00:02, 461.81it/s]
6%|▌ | Scoring GeneralizingEstimator : 73/1225 [00:00<00:02, 425.17it/s]
7%|▋ | Scoring GeneralizingEstimator : 88/1225 [00:00<00:02, 428.57it/s]
9%|▊ | Scoring GeneralizingEstimator : 106/1225 [00:00<00:02, 444.75it/s]
10%|█ | Scoring GeneralizingEstimator : 123/1225 [00:00<00:02, 453.39it/s]
11%|█▏ | Scoring GeneralizingEstimator : 140/1225 [00:00<00:02, 460.10it/s]
13%|█▎ | Scoring GeneralizingEstimator : 157/1225 [00:00<00:02, 464.85it/s]
14%|█▍ | Scoring GeneralizingEstimator : 175/1225 [00:00<00:02, 472.02it/s]
16%|█▌ | Scoring GeneralizingEstimator : 192/1225 [00:00<00:02, 474.94it/s]
17%|█▋ | Scoring GeneralizingEstimator : 209/1225 [00:00<00:02, 477.84it/s]
18%|█▊ | Scoring GeneralizingEstimator : 226/1225 [00:00<00:02, 480.39it/s]
20%|█▉ | Scoring GeneralizingEstimator : 243/1225 [00:00<00:02, 482.15it/s]
21%|██ | Scoring GeneralizingEstimator : 260/1225 [00:00<00:01, 483.90it/s]
23%|██▎ | Scoring GeneralizingEstimator : 278/1225 [00:00<00:01, 487.99it/s]
24%|██▍ | Scoring GeneralizingEstimator : 295/1225 [00:00<00:01, 489.13it/s]
25%|██▌ | Scoring GeneralizingEstimator : 312/1225 [00:00<00:01, 490.07it/s]
27%|██▋ | Scoring GeneralizingEstimator : 329/1225 [00:00<00:01, 490.59it/s]
28%|██▊ | Scoring GeneralizingEstimator : 346/1225 [00:00<00:01, 491.53it/s]
30%|██▉ | Scoring GeneralizingEstimator : 363/1225 [00:00<00:01, 492.47it/s]
31%|███ | Scoring GeneralizingEstimator : 380/1225 [00:00<00:01, 493.20it/s]
32%|███▏ | Scoring GeneralizingEstimator : 397/1225 [00:00<00:01, 493.91it/s]
34%|███▍ | Scoring GeneralizingEstimator : 414/1225 [00:00<00:01, 494.57it/s]
35%|███▌ | Scoring GeneralizingEstimator : 431/1225 [00:00<00:01, 495.15it/s]
37%|███▋ | Scoring GeneralizingEstimator : 448/1225 [00:00<00:01, 495.61it/s]
38%|███▊ | Scoring GeneralizingEstimator : 466/1225 [00:00<00:01, 497.92it/s]
39%|███▉ | Scoring GeneralizingEstimator : 483/1225 [00:00<00:01, 498.28it/s]
41%|████ | Scoring GeneralizingEstimator : 500/1225 [00:01<00:01, 498.66it/s]
42%|████▏ | Scoring GeneralizingEstimator : 517/1225 [00:01<00:01, 498.65it/s]
44%|████▎ | Scoring GeneralizingEstimator : 535/1225 [00:01<00:01, 500.43it/s]
45%|████▌ | Scoring GeneralizingEstimator : 552/1225 [00:01<00:01, 500.66it/s]
46%|████▋ | Scoring GeneralizingEstimator : 569/1225 [00:01<00:01, 500.87it/s]
48%|████▊ | Scoring GeneralizingEstimator : 586/1225 [00:01<00:01, 500.88it/s]
49%|████▉ | Scoring GeneralizingEstimator : 604/1225 [00:01<00:01, 502.69it/s]
51%|█████ | Scoring GeneralizingEstimator : 621/1225 [00:01<00:01, 502.79it/s]
52%|█████▏ | Scoring GeneralizingEstimator : 638/1225 [00:01<00:01, 502.78it/s]
53%|█████▎ | Scoring GeneralizingEstimator : 655/1225 [00:01<00:01, 502.78it/s]
55%|█████▍ | Scoring GeneralizingEstimator : 673/1225 [00:01<00:01, 504.47it/s]
56%|█████▋ | Scoring GeneralizingEstimator : 690/1225 [00:01<00:01, 504.42it/s]
58%|█████▊ | Scoring GeneralizingEstimator : 706/1225 [00:01<00:01, 502.73it/s]
59%|█████▉ | Scoring GeneralizingEstimator : 722/1225 [00:01<00:01, 501.15it/s]
60%|██████ | Scoring GeneralizingEstimator : 739/1225 [00:01<00:00, 501.29it/s]
62%|██████▏ | Scoring GeneralizingEstimator : 757/1225 [00:01<00:00, 502.98it/s]
63%|██████▎ | Scoring GeneralizingEstimator : 772/1225 [00:01<00:00, 499.77it/s]
64%|██████▍ | Scoring GeneralizingEstimator : 783/1225 [00:01<00:00, 490.08it/s]
65%|██████▍ | Scoring GeneralizingEstimator : 794/1225 [00:01<00:00, 480.75it/s]
66%|██████▌ | Scoring GeneralizingEstimator : 806/1225 [00:01<00:00, 473.60it/s]
67%|██████▋ | Scoring GeneralizingEstimator : 823/1225 [00:01<00:00, 475.23it/s]
69%|██████▊ | Scoring GeneralizingEstimator : 841/1225 [00:01<00:00, 478.13it/s]
70%|███████ | Scoring GeneralizingEstimator : 858/1225 [00:01<00:00, 479.29it/s]
71%|███████▏ | Scoring GeneralizingEstimator : 875/1225 [00:01<00:00, 480.52it/s]
73%|███████▎ | Scoring GeneralizingEstimator : 892/1225 [00:01<00:00, 481.71it/s]
74%|███████▍ | Scoring GeneralizingEstimator : 909/1225 [00:01<00:00, 482.62it/s]
76%|███████▌ | Scoring GeneralizingEstimator : 926/1225 [00:01<00:00, 483.61it/s]
77%|███████▋ | Scoring GeneralizingEstimator : 943/1225 [00:01<00:00, 484.50it/s]
78%|███████▊ | Scoring GeneralizingEstimator : 961/1225 [00:01<00:00, 486.70it/s]
80%|███████▉ | Scoring GeneralizingEstimator : 978/1225 [00:02<00:00, 487.54it/s]
81%|████████ | Scoring GeneralizingEstimator : 995/1225 [00:02<00:00, 488.18it/s]
83%|████████▎ | Scoring GeneralizingEstimator : 1012/1225 [00:02<00:00, 488.94it/s]
84%|████████▍ | Scoring GeneralizingEstimator : 1029/1225 [00:02<00:00, 489.36it/s]
85%|████████▌ | Scoring GeneralizingEstimator : 1047/1225 [00:02<00:00, 491.28it/s]
87%|████████▋ | Scoring GeneralizingEstimator : 1064/1225 [00:02<00:00, 491.86it/s]
88%|████████▊ | Scoring GeneralizingEstimator : 1081/1225 [00:02<00:00, 492.46it/s]
90%|████████▉ | Scoring GeneralizingEstimator : 1098/1225 [00:02<00:00, 493.06it/s]
91%|█████████ | Scoring GeneralizingEstimator : 1115/1225 [00:02<00:00, 493.56it/s]
92%|█████████▏| Scoring GeneralizingEstimator : 1132/1225 [00:02<00:00, 493.96it/s]
94%|█████████▍| Scoring GeneralizingEstimator : 1149/1225 [00:02<00:00, 494.44it/s]
95%|█████████▌| Scoring GeneralizingEstimator : 1167/1225 [00:02<00:00, 496.11it/s]
97%|█████████▋| Scoring GeneralizingEstimator : 1184/1225 [00:02<00:00, 496.52it/s]
98%|█████████▊| Scoring GeneralizingEstimator : 1201/1225 [00:02<00:00, 496.90it/s]
99%|█████████▉| Scoring GeneralizingEstimator : 1218/1225 [00:02<00:00, 497.19it/s]
100%|██████████| Scoring GeneralizingEstimator : 1225/1225 [00:02<00:00, 492.11it/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 7.344 seconds)
Estimated memory usage: 129 MB