def test_full_shift_no_shorter_cuts(self, feature_set):
cut_set = make_windowed_cuts_from_features(feature_set=feature_set,
cut_duration=5.0)
assert len(cut_set) == 191
assert len([c for c in cut_set if c.recording_id == 'rec-1']) == 120
assert len([c for c in cut_set if c.recording_id == 'rec-2']) == 71
assert all(c.duration == 5.0 for c in cut_set)
def test_full_shift_with_shorter_cuts(self, feature_set):
cut_set = make_windowed_cuts_from_features(feature_set=feature_set,
cut_duration=5.0,
keep_shorter_windows=True)
assert len(cut_set) == 192
assert len([c for c in cut_set if c.recording_id == 'rec-1']) == 120
assert len([c for c in cut_set if c.recording_id == 'rec-2']) == 72
assert not all(c.duration == 5.0 for c in cut_set)
def test_half_shift_no_shorter_cuts(self, feature_set):
cut_set = make_windowed_cuts_from_features(feature_set=feature_set,
cut_duration=5.0,
cut_shift=2.5)
assert len(cut_set) == 380
# below, the last window is only 2.5s duration
assert len([c for c in cut_set if c.recording_id == 'rec-1']) == 239
# below, the last two windows are 2.0s and 4.5s duration
assert len([c for c in cut_set if c.recording_id == 'rec-2']) == 141
assert all(c.duration == 5.0 for c in cut_set)
def test_half_shift_with_shorter_cuts(self, feature_set):
cut_set = make_windowed_cuts_from_features(
feature_set=feature_set,
cut_duration=5.0,
cut_shift=2.5,
keep_shorter_windows=True,
)
assert len(cut_set) == 383
assert len([c for c in cut_set if c.recording_id == "rec-1"]) == 240
assert len([c for c in cut_set if c.recording_id == "rec-2"]) == 143
assert not all(c.duration == 5.0 for c in cut_set)
def windowed(feature_manifest: Pathlike, output_cut_manifest: Pathlike,
cut_duration: float, cut_shift: Optional[float],
keep_shorter_windows: bool):
"""
Create a CutSet stored in OUTPUT_CUT_MANIFEST from feature regions in FEATURE_MANIFEST.
The feature matrices are traversed in windows with CUT_SHIFT increments, creating cuts of constant CUT_DURATION.
"""
feature_set = FeatureSet.from_json(feature_manifest)
cut_set = make_windowed_cuts_from_features(
feature_set=feature_set,
cut_duration=cut_duration,
cut_shift=cut_shift,
keep_shorter_windows=keep_shorter_windows)
cut_set.to_json(output_cut_manifest)
