def test_wrapper_names_base(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper)
wrap.addFeature("bandPower", 0, bandLimits={"band1":(3,4),
"band2":(10,20)})
featureName="bandPower(0,){'bandLimits': {'band1': (3, 4), 'band2': (10, 20)}}"
self.assertEqual(list(wrap.featuresNames())[0], featureName)
def test_wrapper_names_noargs(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper)
wrap.addFeature("bandPower", 0, bandLimits={"band1":(3,4),
"band2":(10,20)},
hideArgs=True)
featureName="bandPower"
self.assertEqual(list(wrap.featuresNames())[0], featureName)
def test_wrapper_add_features_errors(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::128])
with self.assertRaises(ValueError) as _:
wrap.addFeatures([("synchronizationLikelihood",[(0,3)]),])
with self.assertRaises(AttributeError) as _:
wrap.addFeatures("PFD")
def test_branches(self):
helpers.CSVHelper(self.test_file_name,
sampleRate=128,
ICA=True,
lowpass=24,
highpass=2,
normalize=True)
h = helpers.Helper(self.test_data)
len(h)
def test_wrapper_segmentation(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::128], label="test",
segmentation=[(("0","2"),"l1"),
((384,512),"l2")])
wrap.addFeature("bandPower", 0)
label = wrap.getAllFeatures()["segment_label"]
self.assertEqual(label[0], "l1")
self.assertEqual(label[1], "l1")
self.assertEqual(label[3], "l2")
def test_channels_names(self):
h = helpers.CSVHelper(self.test_file_name,
windowSize=64,
names=["F1", "F2", "F3", "F4", "F5"],
selectedSignals=["F1", "F3"])
self.assertListEqual(h.getNames(), ["F1", "F3"])
self.assertListEqual(h.getNames([1]), ["F3"])
# Not sure if this should be allowed, but it will be keeped by the
# moment.
h.selectSignals([1, 0, "F1"])
self.assertListEqual(h.getNames(), ["F3", "F1", "F1"])
def test_custom_features_params(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::128])
wrap.addCustomFeature(lambda x, y: np.linalg.norm(x-y, ord=2),
channels=[0,1,2], twoChannels=True,
name="norm")
f1 = wrap.getAllFeatures()
wrap = wrapper.Wrapper(helper[::128])
wrap.addCustomFeature(lambda x, y, *args, **kwargs: np.linalg.norm(x-y, *args, **kwargs),
channels=[0,1,2], twoChannels=True,
name="norm" ,customKwargs={"ord":2})
f2 = wrap.getAllFeatures()
pd.testing.assert_frame_equal(f1, f2, check_names = False)
def test_wrapper_all_features_progress(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::16], showProgress=True,
segmentation=[(("0","2"), "l1"),
(("6","10"), "l2")],
onlySegments=True)
wrap.addFeature("bandPower", 0)
wrap.addFeature.DFA([1,2])
wrap.addFeature.HFD(kMax=8)
wrap.addFeatures([("synchronizationLikelihood",[],{}),
"PFD"])
wrap.addCustomFeature(np.mean)
wrap.addCustomFeature(lambda x, y: np.linalg.norm(x-y),
channels=[0,1,2], twoChannels=True)
features = wrap.getAllFeatures()
def test_wrapper_segmentation_only_segments(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::128], label="test",
segmentation=[(("0","2"),"l1"),
((384,512),"l2"),
(("6","8"), "l3")],
onlySegments = True)
wrap.addFeature("bandPower", 0)
features = wrap.getAllFeatures()
label = features["segment_label"]
self.assertEqual(label[0], "l1")
self.assertEqual(label[1], "l1")
self.assertEqual(label[3], "l2")
self.assertEqual(label[6], "l3")
self.assertEqual(label[7], "l3")
def test_wrapper_all_features_stored(self):
helper = helpers.CSVHelper(self.test_file_name, sampleRate=128)
wrap = wrapper.Wrapper(helper[::128])
wrap.addFeature("bandPower", 0)
wrap.addFeature.DFA([1,2])
wrap.addFeature.HFD(kMax=8)
wrap.addFeatures([("synchronizationLikelihood",[(0,3)],{}),
"PFD"])
wrap.addCustomFeature(np.mean)
wrap.addCustomFeature(lambda x, y: np.linalg.norm(x-y),
channels=[0,1,2], twoChannels=True)
features = wrap.getAllFeatures()
stored = wrap.getStoredFeatures()
pd.testing.assert_frame_equal(features, stored)
pd.testing.assert_series_equal(features.iloc[-1], wrap.getFeatures().drop(wrap._indexName),
check_names=False)
def test_CSVHelperNames(self):
h = helpers.CSVHelper(self.test_file_name, sampleRate=128)
self.assertListEqual(h.getNames(), ["C1", "C2", "C3", "C4", "C5"])
h = helpers.CSVHelper(self.test_file_noname, sampleRate=128)
self.assertListEqual(h.getNames(), ["0", "1", "2", "3", "4"])
from eeglib import wrapper, helpers
helper = helpers.CSVHelper("fake_EEG_signal.csv", windowSize=128)
wrap = wrapper.Wrapper(helper)
wrap.addFeature.HFD()
wrap.addFeature.DFT()
wrap.addFeature.synchronizationLikelihood()
features = wrap.getAllFeatures()