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
raw_fname = data_path + '/MEG/sample/sample_audvis_filt-0-40_raw.fif'
events_fname = data_path + '/MEG/sample/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')

Out:

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)

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='lbfgs'))
time_gen = GeneralizingEstimator(clf, scoring='roc_auc', n_jobs=1,
                                 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)

Out:

  0%|          | Fitting GeneralizingEstimator : 0/35 [00:00<?,       ?it/s]
  3%|2         | Fitting GeneralizingEstimator : 1/35 [00:00<00:01,   29.21it/s]
  6%|5         | Fitting GeneralizingEstimator : 2/35 [00:00<00:01,   28.70it/s]
  9%|8         | Fitting GeneralizingEstimator : 3/35 [00:00<00:01,   20.18it/s]
 14%|#4        | Fitting GeneralizingEstimator : 5/35 [00:00<00:01,   27.98it/s]
 20%|##        | Fitting GeneralizingEstimator : 7/35 [00:00<00:01,   24.97it/s]
 23%|##2       | Fitting GeneralizingEstimator : 8/35 [00:00<00:01,   22.85it/s]
 26%|##5       | Fitting GeneralizingEstimator : 9/35 [00:00<00:01,   23.51it/s]
 29%|##8       | Fitting GeneralizingEstimator : 10/35 [00:00<00:01,   22.07it/s]
 34%|###4      | Fitting GeneralizingEstimator : 12/35 [00:00<00:00,   25.18it/s]
 37%|###7      | Fitting GeneralizingEstimator : 13/35 [00:00<00:00,   25.54it/s]
 40%|####      | Fitting GeneralizingEstimator : 14/35 [00:00<00:00,   23.90it/s]
 43%|####2     | Fitting GeneralizingEstimator : 15/35 [00:00<00:00,   24.30it/s]
 49%|####8     | Fitting GeneralizingEstimator : 17/35 [00:00<00:00,   24.78it/s]
 51%|#####1    | Fitting GeneralizingEstimator : 18/35 [00:00<00:00,   22.18it/s]
 54%|#####4    | Fitting GeneralizingEstimator : 19/35 [00:00<00:00,   21.85it/s]
 60%|######    | Fitting GeneralizingEstimator : 21/35 [00:00<00:00,   23.86it/s]
 63%|######2   | Fitting GeneralizingEstimator : 22/35 [00:00<00:00,   23.01it/s]
 66%|######5   | Fitting GeneralizingEstimator : 23/35 [00:00<00:00,   23.30it/s]
 69%|######8   | Fitting GeneralizingEstimator : 24/35 [00:01<00:00,   22.56it/s]
 74%|#######4  | Fitting GeneralizingEstimator : 26/35 [00:01<00:00,   24.35it/s]
 77%|#######7  | Fitting GeneralizingEstimator : 27/35 [00:01<00:00,   22.38it/s]
 80%|########  | Fitting GeneralizingEstimator : 28/35 [00:01<00:00,   21.84it/s]
 83%|########2 | Fitting GeneralizingEstimator : 29/35 [00:01<00:00,   22.00it/s]
 86%|########5 | Fitting GeneralizingEstimator : 30/35 [00:01<00:00,   21.60it/s]
 89%|########8 | Fitting GeneralizingEstimator : 31/35 [00:01<00:00,   21.85it/s]
 91%|#########1| Fitting GeneralizingEstimator : 32/35 [00:01<00:00,   21.40it/s]
 97%|#########7| Fitting GeneralizingEstimator : 34/35 [00:01<00:00,   22.65it/s]
100%|##########| Fitting GeneralizingEstimator : 35/35 [00:01<00:00,   21.88it/s]
100%|##########| Fitting GeneralizingEstimator : 35/35 [00:01<00:00,   22.36it/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)

Out:

  0%|          | Scoring GeneralizingEstimator : 0/1225 [00:00<?,       ?it/s]
  1%|1         | Scoring GeneralizingEstimator : 15/1225 [00:00<00:02,  437.11it/s]
  3%|3         | Scoring GeneralizingEstimator : 37/1225 [00:00<00:02,  544.09it/s]
  3%|3         | Scoring GeneralizingEstimator : 38/1225 [00:00<00:03,  321.01it/s]
  5%|4         | Scoring GeneralizingEstimator : 56/1225 [00:00<00:03,  372.16it/s]
  5%|5         | Scoring GeneralizingEstimator : 67/1225 [00:00<00:03,  303.34it/s]
  7%|7         | Scoring GeneralizingEstimator : 86/1225 [00:00<00:03,  342.28it/s]
  8%|7         | Scoring GeneralizingEstimator : 96/1225 [00:00<00:03,  297.37it/s]
  9%|9         | Scoring GeneralizingEstimator : 114/1225 [00:00<00:03,  323.85it/s]
 10%|#         | Scoring GeneralizingEstimator : 124/1225 [00:00<00:03,  291.20it/s]
 12%|#1        | Scoring GeneralizingEstimator : 143/1225 [00:00<00:03,  316.17it/s]
 13%|#2        | Scoring GeneralizingEstimator : 154/1225 [00:00<00:03,  292.24it/s]
 14%|#4        | Scoring GeneralizingEstimator : 174/1225 [00:00<00:03,  315.89it/s]
 15%|#5        | Scoring GeneralizingEstimator : 185/1225 [00:00<00:03,  294.97it/s]
 17%|#6        | Scoring GeneralizingEstimator : 205/1225 [00:00<00:03,  315.64it/s]
 18%|#7        | Scoring GeneralizingEstimator : 215/1225 [00:00<00:03,  294.94it/s]
 19%|#8        | Scoring GeneralizingEstimator : 229/1225 [00:00<00:03,  302.37it/s]
 19%|#9        | Scoring GeneralizingEstimator : 236/1225 [00:00<00:03,  279.73it/s]
 21%|##        | Scoring GeneralizingEstimator : 256/1225 [00:00<00:03,  297.89it/s]
 22%|##1       | Scoring GeneralizingEstimator : 267/1225 [00:00<00:03,  283.82it/s]
 23%|##3       | Scoring GeneralizingEstimator : 287/1225 [00:00<00:03,  300.48it/s]
 24%|##4       | Scoring GeneralizingEstimator : 297/1225 [00:01<00:03,  285.53it/s]
 26%|##5       | Scoring GeneralizingEstimator : 315/1225 [00:01<00:03,  298.20it/s]
 27%|##6       | Scoring GeneralizingEstimator : 325/1225 [00:01<00:03,  298.00it/s]
 28%|##7       | Scoring GeneralizingEstimator : 340/1225 [00:01<00:03,  290.43it/s]
 28%|##8       | Scoring GeneralizingEstimator : 346/1225 [00:01<00:03,  272.88it/s]
 30%|##9       | Scoring GeneralizingEstimator : 365/1225 [00:01<00:03,  286.44it/s]
 31%|###       | Scoring GeneralizingEstimator : 375/1225 [00:01<00:03,  274.52it/s]
 32%|###1      | Scoring GeneralizingEstimator : 389/1225 [00:01<00:02,  280.79it/s]
 32%|###2      | Scoring GeneralizingEstimator : 397/1225 [00:01<00:03,  267.16it/s]
 34%|###3      | Scoring GeneralizingEstimator : 411/1225 [00:01<00:02,  273.56it/s]
 34%|###4      | Scoring GeneralizingEstimator : 417/1225 [00:01<00:03,  258.42it/s]
 35%|###5      | Scoring GeneralizingEstimator : 429/1225 [00:01<00:03,  262.46it/s]
 36%|###5      | Scoring GeneralizingEstimator : 436/1225 [00:01<00:03,  249.84it/s]
 37%|###6      | Scoring GeneralizingEstimator : 448/1225 [00:01<00:03,  253.93it/s]
 37%|###7      | Scoring GeneralizingEstimator : 454/1225 [00:01<00:03,  241.30it/s]
 38%|###8      | Scoring GeneralizingEstimator : 470/1225 [00:01<00:03,  250.65it/s]
 39%|###9      | Scoring GeneralizingEstimator : 478/1225 [00:01<00:03,  240.68it/s]
 41%|####      | Scoring GeneralizingEstimator : 498/1225 [00:01<00:02,  254.67it/s]
 42%|####1     | Scoring GeneralizingEstimator : 509/1225 [00:01<00:02,  248.00it/s]
 43%|####3     | Scoring GeneralizingEstimator : 527/1225 [00:01<00:02,  259.14it/s]
 44%|####3     | Scoring GeneralizingEstimator : 538/1225 [00:02<00:02,  252.21it/s]
 45%|####5     | Scoring GeneralizingEstimator : 557/1225 [00:02<00:02,  264.15it/s]
 46%|####6     | Scoring GeneralizingEstimator : 569/1225 [00:02<00:02,  258.14it/s]
 48%|####8     | Scoring GeneralizingEstimator : 589/1225 [00:02<00:02,  270.87it/s]
 49%|####8     | Scoring GeneralizingEstimator : 600/1225 [00:02<00:02,  263.21it/s]
 51%|#####     | Scoring GeneralizingEstimator : 619/1225 [00:02<00:02,  274.45it/s]
 51%|#####1    | Scoring GeneralizingEstimator : 630/1225 [00:02<00:02,  266.69it/s]
 53%|#####3    | Scoring GeneralizingEstimator : 650/1225 [00:02<00:02,  278.77it/s]
 54%|#####3    | Scoring GeneralizingEstimator : 661/1225 [00:02<00:02,  270.81it/s]
 56%|#####5    | Scoring GeneralizingEstimator : 680/1225 [00:02<00:01,  281.57it/s]
 56%|#####6    | Scoring GeneralizingEstimator : 691/1225 [00:02<00:01,  273.41it/s]
 58%|#####7    | Scoring GeneralizingEstimator : 710/1225 [00:02<00:01,  283.95it/s]
 59%|#####8    | Scoring GeneralizingEstimator : 720/1225 [00:02<00:01,  274.00it/s]
 60%|######    | Scoring GeneralizingEstimator : 741/1225 [00:02<00:01,  286.59it/s]
 61%|######1   | Scoring GeneralizingEstimator : 750/1225 [00:02<00:01,  276.79it/s]
 63%|######2   | Scoring GeneralizingEstimator : 769/1225 [00:02<00:01,  286.88it/s]
 64%|######3   | Scoring GeneralizingEstimator : 779/1225 [00:02<00:01,  277.70it/s]
 65%|######5   | Scoring GeneralizingEstimator : 799/1225 [00:02<00:01,  288.94it/s]
 66%|######6   | Scoring GeneralizingEstimator : 809/1225 [00:02<00:01,  279.51it/s]
 68%|######7   | Scoring GeneralizingEstimator : 830/1225 [00:02<00:01,  291.72it/s]
 69%|######8   | Scoring GeneralizingEstimator : 842/1225 [00:03<00:01,  284.13it/s]
 70%|#######   | Scoring GeneralizingEstimator : 862/1225 [00:03<00:01,  295.04it/s]
 71%|#######1  | Scoring GeneralizingEstimator : 872/1225 [00:03<00:01,  295.06it/s]
 73%|#######2  | Scoring GeneralizingEstimator : 893/1225 [00:03<00:01,  296.10it/s]
 74%|#######3  | Scoring GeneralizingEstimator : 903/1225 [00:03<00:01,  286.81it/s]
 75%|#######4  | Scoring GeneralizingEstimator : 917/1225 [00:03<00:01,  291.21it/s]
 75%|#######5  | Scoring GeneralizingEstimator : 924/1225 [00:03<00:01,  278.69it/s]
 77%|#######7  | Scoring GeneralizingEstimator : 944/1225 [00:03<00:00,  289.64it/s]
 78%|#######7  | Scoring GeneralizingEstimator : 955/1225 [00:03<00:00,  281.34it/s]
 80%|#######9  | Scoring GeneralizingEstimator : 974/1225 [00:03<00:00,  291.14it/s]
 80%|########  | Scoring GeneralizingEstimator : 985/1225 [00:03<00:00,  282.77it/s]
 81%|########1 | Scoring GeneralizingEstimator : 997/1225 [00:03<00:00,  285.17it/s]
 82%|########1 | Scoring GeneralizingEstimator : 1002/1225 [00:03<00:00,  280.28it/s]
 83%|########3 | Scoring GeneralizingEstimator : 1019/1225 [00:03<00:00,  278.31it/s]
 84%|########3 | Scoring GeneralizingEstimator : 1028/1225 [00:03<00:00,  269.29it/s]
 86%|########5 | Scoring GeneralizingEstimator : 1049/1225 [00:03<00:00,  281.48it/s]
 87%|########6 | Scoring GeneralizingEstimator : 1060/1225 [00:03<00:00,  273.81it/s]
 88%|########8 | Scoring GeneralizingEstimator : 1080/1225 [00:03<00:00,  284.77it/s]
 89%|########9 | Scoring GeneralizingEstimator : 1091/1225 [00:03<00:00,  276.93it/s]
 91%|######### | Scoring GeneralizingEstimator : 1111/1225 [00:03<00:00,  287.79it/s]
 92%|#########1| Scoring GeneralizingEstimator : 1123/1225 [00:04<00:00,  280.72it/s]
 93%|#########2| Scoring GeneralizingEstimator : 1137/1225 [00:04<00:00,  285.30it/s]
 93%|#########3| Scoring GeneralizingEstimator : 1144/1225 [00:04<00:00,  273.40it/s]
 95%|#########4| Scoring GeneralizingEstimator : 1163/1225 [00:04<00:00,  283.33it/s]
 96%|#########5| Scoring GeneralizingEstimator : 1172/1225 [00:04<00:00,  273.61it/s]
 97%|#########7| Scoring GeneralizingEstimator : 1191/1225 [00:04<00:00,  283.52it/s]
 98%|#########8| Scoring GeneralizingEstimator : 1201/1225 [00:04<00:00,  274.74it/s]
100%|#########9| Scoring GeneralizingEstimator : 1220/1225 [00:04<00:00,  284.59it/s]
100%|##########| Scoring GeneralizingEstimator : 1225/1225 [00:04<00:00,  281.24it/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()
Generalization across time and condition

References

1

Jean-Rémi King and Stanislas Dehaene. Characterizing the dynamics of mental representations: the temporal generalization method. Trends in Cognitive Sciences, 18(4):203–210, 2014. doi:10.1016/j.tics.2014.01.002.

Total running time of the script: ( 0 minutes 11.512 seconds)

Estimated memory usage: 128 MB

Gallery generated by Sphinx-Gallery