def test_trialwise_predict_and_predict_proba():
preds = np.array([
[0.125, 0.875],
[1., 0.],
[0.8, 0.2],
[0.9, 0.1],
])
clf = EEGClassifier(MockModule(preds), optimizer=optim.Adam, batch_size=32)
clf.initialize()
np.testing.assert_array_equal(preds.argmax(1), clf.predict(MockDataset()))
np.testing.assert_array_equal(preds, clf.predict_proba(MockDataset()))
def test_cropped_predict_and_predict_proba_not_aggregate_predictions():
preds = np.array([
[[0.2, 0.1, 0.1, 0.1], [0.8, 0.9, 0.9, 0.9]],
[[1.0, 1.0, 1.0, 1.0], [0.0, 0.0, 0.0, 0.0]],
[[1.0, 1.0, 1.0, 0.2], [0.0, 0.0, 0.0, 0.8]],
[[0.9, 0.8, 0.9, 1.0], [0.1, 0.2, 0.1, 0.0]],
])
clf = EEGClassifier(MockModule(preds),
cropped=True,
criterion=CroppedLoss,
criterion__loss_function=nll_loss,
optimizer=optim.Adam,
batch_size=32,
aggregate_predictions=False)
clf.initialize()
np.testing.assert_array_equal(preds.argmax(1), clf.predict(MockDataset()))
np.testing.assert_array_equal(preds, clf.predict_proba(MockDataset()))
def test_cropped_predict_and_predict_proba():
preds = np.array([
[[0.2, 0.1, 0.1, 0.1], [0.8, 0.9, 0.9, 0.9]],
[[1.0, 1.0, 1.0, 1.0], [0.0, 0.0, 0.0, 0.0]],
[[1.0, 1.0, 1.0, 0.2], [0.0, 0.0, 0.0, 0.8]],
[[0.9, 0.8, 0.9, 1.0], [0.1, 0.2, 0.1, 0.0]],
])
clf = EEGClassifier(MockModule(preds),
cropped=True,
criterion=CroppedLoss,
criterion__loss_function=nll_loss,
optimizer=optim.Adam,
batch_size=32)
clf.initialize()
# for cropped decoding classifier returns one label for each trial (averaged over all crops)
np.testing.assert_array_equal(
preds.mean(-1).argmax(1), clf.predict(MockDataset()))
# for cropped decoding classifier returns values for each trial (average over all crops)
np.testing.assert_array_equal(preds.mean(-1),
clf.predict_proba(MockDataset()))
