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
import matplotlib.pyplot as plt
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
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., 30., 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 sec)
[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.
[Parallel(n_jobs=1)]: Done 1 out of 1 | elapsed: 0.0s remaining: 0.0s
[Parallel(n_jobs=1)]: Done 2 out of 2 | elapsed: 0.0s remaining: 0.0s
[Parallel(n_jobs=1)]: Done 3 out of 3 | elapsed: 0.0s remaining: 0.0s
[Parallel(n_jobs=1)]: Done 4 out of 4 | elapsed: 0.0s remaining: 0.0s
[Parallel(n_jobs=1)]: Done 366 out of 366 | elapsed: 0.7s finished
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(),
y=epochs['Left'].events[:, 2] > 2)
0%| | Fitting GeneralizingEstimator : 0/35 [00:00<?, ?it/s]
6%|5 | Fitting GeneralizingEstimator : 2/35 [00:00<00:00, 58.21it/s]
11%|#1 | Fitting GeneralizingEstimator : 4/35 [00:00<00:00, 58.69it/s]
17%|#7 | Fitting GeneralizingEstimator : 6/35 [00:00<00:00, 58.84it/s]
23%|##2 | Fitting GeneralizingEstimator : 8/35 [00:00<00:00, 58.92it/s]
31%|###1 | Fitting GeneralizingEstimator : 11/35 [00:00<00:00, 65.46it/s]
40%|#### | Fitting GeneralizingEstimator : 14/35 [00:00<00:00, 69.87it/s]
49%|####8 | Fitting GeneralizingEstimator : 17/35 [00:00<00:00, 72.99it/s]
54%|#####4 | Fitting GeneralizingEstimator : 19/35 [00:00<00:00, 70.95it/s]
63%|######2 | Fitting GeneralizingEstimator : 22/35 [00:00<00:00, 73.36it/s]
71%|#######1 | Fitting GeneralizingEstimator : 25/35 [00:00<00:00, 75.17it/s]
77%|#######7 | Fitting GeneralizingEstimator : 27/35 [00:00<00:00, 73.33it/s]
86%|########5 | Fitting GeneralizingEstimator : 30/35 [00:00<00:00, 74.91it/s]
91%|#########1| Fitting GeneralizingEstimator : 32/35 [00:00<00:00, 73.28it/s]
100%|##########| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 75.35it/s]
100%|##########| Fitting GeneralizingEstimator : 35/35 [00:00<00:00, 74.08it/s]
Score on the epochs where the stimulus was presented to the right.
scores = time_gen.score(X=epochs['Right'].get_data(),
y=epochs['Right'].events[:, 2] > 2)
0%| | Scoring GeneralizingEstimator : 0/1225 [00:00<?, ?it/s]
1%|1 | Scoring GeneralizingEstimator : 14/1225 [00:00<00:02, 409.21it/s]
3%|2 | Scoring GeneralizingEstimator : 32/1225 [00:00<00:02, 471.58it/s]
4%|4 | Scoring GeneralizingEstimator : 50/1225 [00:00<00:02, 493.03it/s]
6%|5 | Scoring GeneralizingEstimator : 69/1225 [00:00<00:02, 511.93it/s]
7%|7 | Scoring GeneralizingEstimator : 89/1225 [00:00<00:02, 529.58it/s]
9%|8 | Scoring GeneralizingEstimator : 109/1225 [00:00<00:02, 541.42it/s]
11%|# | Scoring GeneralizingEstimator : 129/1225 [00:00<00:01, 549.96it/s]
12%|#1 | Scoring GeneralizingEstimator : 145/1225 [00:00<00:02, 538.48it/s]
14%|#3 | Scoring GeneralizingEstimator : 166/1225 [00:00<00:01, 549.50it/s]
15%|#5 | Scoring GeneralizingEstimator : 186/1225 [00:00<00:01, 554.60it/s]
17%|#6 | Scoring GeneralizingEstimator : 206/1225 [00:00<00:01, 558.57it/s]
19%|#8 | Scoring GeneralizingEstimator : 227/1225 [00:00<00:01, 565.41it/s]
20%|#9 | Scoring GeneralizingEstimator : 241/1225 [00:00<00:01, 549.67it/s]
21%|##1 | Scoring GeneralizingEstimator : 260/1225 [00:00<00:01, 550.65it/s]
23%|##2 | Scoring GeneralizingEstimator : 280/1225 [00:00<00:01, 554.44it/s]
24%|##4 | Scoring GeneralizingEstimator : 299/1225 [00:00<00:01, 554.79it/s]
26%|##5 | Scoring GeneralizingEstimator : 318/1225 [00:00<00:01, 555.53it/s]
27%|##7 | Scoring GeneralizingEstimator : 331/1225 [00:00<00:01, 541.41it/s]
28%|##8 | Scoring GeneralizingEstimator : 347/1225 [00:00<00:01, 535.97it/s]
29%|##9 | Scoring GeneralizingEstimator : 359/1225 [00:00<00:01, 521.68it/s]
30%|### | Scoring GeneralizingEstimator : 373/1225 [00:00<00:01, 513.12it/s]
32%|###1 | Scoring GeneralizingEstimator : 386/1225 [00:00<00:01, 503.62it/s]
33%|###2 | Scoring GeneralizingEstimator : 400/1225 [00:00<00:01, 496.96it/s]
34%|###3 | Scoring GeneralizingEstimator : 415/1225 [00:00<00:01, 493.01it/s]
35%|###4 | Scoring GeneralizingEstimator : 427/1225 [00:00<00:01, 483.40it/s]
36%|###5 | Scoring GeneralizingEstimator : 439/1225 [00:00<00:01, 474.62it/s]
37%|###6 | Scoring GeneralizingEstimator : 452/1225 [00:00<00:01, 468.57it/s]
38%|###7 | Scoring GeneralizingEstimator : 465/1225 [00:00<00:01, 463.06it/s]
39%|###9 | Scoring GeneralizingEstimator : 478/1225 [00:00<00:01, 458.04it/s]
40%|#### | Scoring GeneralizingEstimator : 490/1225 [00:01<00:01, 451.53it/s]
41%|#### | Scoring GeneralizingEstimator : 501/1225 [00:01<00:01, 443.63it/s]
42%|####2 | Scoring GeneralizingEstimator : 515/1225 [00:01<00:01, 441.66it/s]
43%|####2 | Scoring GeneralizingEstimator : 524/1225 [00:01<00:01, 430.77it/s]
44%|####3 | Scoring GeneralizingEstimator : 537/1225 [00:01<00:01, 427.85it/s]
45%|####4 | Scoring GeneralizingEstimator : 548/1225 [00:01<00:01, 421.63it/s]
46%|####5 | Scoring GeneralizingEstimator : 559/1225 [00:01<00:01, 415.86it/s]
47%|####6 | Scoring GeneralizingEstimator : 573/1225 [00:01<00:01, 415.48it/s]
48%|####8 | Scoring GeneralizingEstimator : 588/1225 [00:01<00:01, 415.66it/s]
49%|####8 | Scoring GeneralizingEstimator : 600/1225 [00:01<00:01, 412.14it/s]
50%|##### | Scoring GeneralizingEstimator : 614/1225 [00:01<00:01, 412.22it/s]
51%|#####1 | Scoring GeneralizingEstimator : 626/1225 [00:01<00:01, 408.40it/s]
52%|#####2 | Scoring GeneralizingEstimator : 638/1225 [00:01<00:01, 405.36it/s]
53%|#####3 | Scoring GeneralizingEstimator : 653/1225 [00:01<00:01, 407.34it/s]
55%|#####4 | Scoring GeneralizingEstimator : 668/1225 [00:01<00:01, 409.14it/s]
56%|#####5 | Scoring GeneralizingEstimator : 681/1225 [00:01<00:01, 407.64it/s]
57%|#####6 | Scoring GeneralizingEstimator : 694/1225 [00:01<00:01, 406.35it/s]
58%|#####7 | Scoring GeneralizingEstimator : 706/1225 [00:01<00:01, 403.52it/s]
59%|#####8 | Scoring GeneralizingEstimator : 719/1225 [00:01<00:01, 402.46it/s]
60%|#####9 | Scoring GeneralizingEstimator : 731/1225 [00:01<00:01, 399.67it/s]
61%|###### | Scoring GeneralizingEstimator : 746/1225 [00:01<00:01, 401.99it/s]
62%|######1 | Scoring GeneralizingEstimator : 756/1225 [00:01<00:01, 396.23it/s]
63%|######2 | Scoring GeneralizingEstimator : 768/1225 [00:01<00:01, 394.04it/s]
64%|######3 | Scoring GeneralizingEstimator : 780/1225 [00:01<00:01, 391.96it/s]
65%|######4 | Scoring GeneralizingEstimator : 792/1225 [00:01<00:01, 389.95it/s]
66%|######5 | Scoring GeneralizingEstimator : 804/1225 [00:01<00:01, 388.06it/s]
67%|######6 | Scoring GeneralizingEstimator : 817/1225 [00:01<00:01, 387.83it/s]
68%|######7 | Scoring GeneralizingEstimator : 831/1225 [00:01<00:01, 389.24it/s]
69%|######9 | Scoring GeneralizingEstimator : 846/1225 [00:01<00:00, 392.07it/s]
70%|####### | Scoring GeneralizingEstimator : 859/1225 [00:02<00:00, 391.44it/s]
71%|#######1 | Scoring GeneralizingEstimator : 871/1225 [00:02<00:00, 389.39it/s]
72%|#######2 | Scoring GeneralizingEstimator : 885/1225 [00:02<00:00, 390.52it/s]
73%|#######3 | Scoring GeneralizingEstimator : 899/1225 [00:02<00:00, 391.63it/s]
74%|#######4 | Scoring GeneralizingEstimator : 911/1225 [00:02<00:00, 389.71it/s]
76%|#######5 | Scoring GeneralizingEstimator : 925/1225 [00:02<00:00, 390.93it/s]
77%|#######6 | Scoring GeneralizingEstimator : 942/1225 [00:02<00:00, 396.67it/s]
78%|#######8 | Scoring GeneralizingEstimator : 956/1225 [00:02<00:00, 397.18it/s]
79%|#######9 | Scoring GeneralizingEstimator : 970/1225 [00:02<00:00, 397.95it/s]
80%|######## | Scoring GeneralizingEstimator : 983/1225 [00:02<00:00, 397.00it/s]
81%|########1 | Scoring GeneralizingEstimator : 996/1225 [00:02<00:00, 396.39it/s]
83%|########2 | Scoring GeneralizingEstimator : 1014/1225 [00:02<00:00, 403.27it/s]
84%|########3 | Scoring GeneralizingEstimator : 1025/1225 [00:02<00:00, 399.19it/s]
85%|########4 | Scoring GeneralizingEstimator : 1036/1225 [00:02<00:00, 395.46it/s]
86%|########5 | Scoring GeneralizingEstimator : 1051/1225 [00:02<00:00, 397.87it/s]
87%|########6 | Scoring GeneralizingEstimator : 1065/1225 [00:02<00:00, 398.70it/s]
88%|########8 | Scoring GeneralizingEstimator : 1078/1225 [00:02<00:00, 397.73it/s]
89%|########8 | Scoring GeneralizingEstimator : 1090/1225 [00:02<00:00, 395.54it/s]
90%|########9 | Scoring GeneralizingEstimator : 1101/1225 [00:02<00:00, 391.94it/s]
91%|######### | Scoring GeneralizingEstimator : 1113/1225 [00:02<00:00, 390.05it/s]
92%|#########1| Scoring GeneralizingEstimator : 1126/1225 [00:02<00:00, 389.72it/s]
93%|#########3| Scoring GeneralizingEstimator : 1141/1225 [00:02<00:00, 391.81it/s]
94%|#########4| Scoring GeneralizingEstimator : 1152/1225 [00:02<00:00, 388.48it/s]
95%|#########5| Scoring GeneralizingEstimator : 1164/1225 [00:02<00:00, 386.57it/s]
96%|#########6| Scoring GeneralizingEstimator : 1176/1225 [00:02<00:00, 384.75it/s]
97%|#########6| Scoring GeneralizingEstimator : 1188/1225 [00:02<00:00, 383.05it/s]
98%|#########8| Scoring GeneralizingEstimator : 1201/1225 [00:02<00:00, 383.13it/s]
99%|#########8| Scoring GeneralizingEstimator : 1212/1225 [00:02<00:00, 380.15it/s]
100%|#########9| Scoring GeneralizingEstimator : 1224/1225 [00:02<00:00, 378.81it/s]
100%|##########| Scoring GeneralizingEstimator : 1225/1225 [00:02<00:00, 414.40it/s]
Plot
fig, ax = plt.subplots(1)
im = ax.matshow(scores, vmin=0, vmax=1., cmap='RdBu_r', origin='lower',
extent=epochs.times[[0, -1, 0, -1]])
ax.axhline(0., color='k')
ax.axvline(0., color='k')
ax.xaxis.set_ticks_position('bottom')
ax.set_xlabel('Testing Time (s)')
ax.set_ylabel('Training Time (s)')
ax.set_title('Generalization across time and condition')
plt.colorbar(im, ax=ax)
plt.show()
References#
Total running time of the script: ( 0 minutes 8.033 seconds)
Estimated memory usage: 128 MB