aiexperiments-ai-duet/server/magenta/music/melodies_lib_test.py
Yotam Mann ff837cec16 server
2016-11-11 13:53:51 -05:00

714 lines
31 KiB
Python

# Copyright 2016 Google Inc. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Tests for melodies_lib."""
# internal imports
import tensorflow as tf
from magenta.common import sequence_example_lib
from magenta.music import constants
from magenta.music import melodies_lib
from magenta.music import sequences_lib
from magenta.music import testing_lib
NUM_SPECIAL_EVENTS = constants.NUM_SPECIAL_MELODY_EVENTS
NOTE_OFF = constants.MELODY_NOTE_OFF
NO_EVENT = constants.MELODY_NO_EVENT
class MelodiesLibTest(tf.test.TestCase):
def setUp(self):
self.quantized_sequence = sequences_lib.QuantizedSequence()
self.quantized_sequence.qpm = 60.0
self.quantized_sequence.steps_per_quarter = 4
def testGetNoteHistogram(self):
events = [NO_EVENT, NOTE_OFF, 12 * 2 + 1, 12 * 3, 12 * 5 + 11, 12 * 6 + 3,
12 * 4 + 11]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
expected = [1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2]
self.assertEqual(expected, list(melody.get_note_histogram()))
events = [0, 1, NO_EVENT, NOTE_OFF, 12 * 2 + 1, 12 * 3, 12 * 6 + 3,
12 * 5 + 11, NO_EVENT, 12 * 4 + 11, 12 * 7 + 1]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
expected = [2, 3, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2]
self.assertEqual(expected, list(melody.get_note_histogram()))
melody = melodies_lib.MonophonicMelody()
expected = [0] * 12
self.assertEqual(expected, list(melody.get_note_histogram()))
def testGetKeyHistogram(self):
# One C.
events = [NO_EVENT, 12 * 5, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
expected = [1, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0]
self.assertListEqual(expected, list(melody.get_major_key_histogram()))
# One C and one C#.
events = [NO_EVENT, 12 * 5, NOTE_OFF, 12 * 7 + 1, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
expected = [1, 2, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1]
self.assertListEqual(expected, list(melody.get_major_key_histogram()))
# One C, one C#, and one D.
events = [NO_EVENT, 12 * 5, NOTE_OFF, 12 * 7 + 1, NO_EVENT, 12 * 9 + 2]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
expected = [2, 2, 2, 2, 1, 2, 1, 2, 2, 2, 2, 1]
self.assertListEqual(expected, list(melody.get_major_key_histogram()))
def testGetMajorKey(self):
# D Major.
events = [NO_EVENT, 12 * 2 + 2, 12 * 3 + 4, 12 * 5 + 1, 12 * 6 + 6,
12 * 4 + 11, 12 * 3 + 9, 12 * 5 + 7, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
self.assertEqual(2, melody.get_major_key())
# C# Major with accidentals.
events = [NO_EVENT, 12 * 2 + 1, 12 * 4 + 8, 12 * 5 + 5, 12 * 6 + 6,
12 * 3 + 3, 12 * 2 + 11, 12 * 3 + 10, 12 * 5, 12 * 2 + 8,
12 * 4 + 1, 12 * 3 + 5, 12 * 5 + 9, 12 * 4 + 3, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
self.assertEqual(1, melody.get_major_key())
# One note in C Major.
events = [NO_EVENT, 12 * 2 + 11, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
self.assertEqual(0, melody.get_major_key())
def testTranspose(self):
# MonophonicMelody transposed down 5 half steps. 2 octave range.
events = [12 * 5 + 4, NO_EVENT, 12 * 5 + 5, NOTE_OFF, 12 * 6, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.transpose(transpose_amount=-5, min_note=12 * 5, max_note=12 * 7)
expected = [12 * 5 + 11, NO_EVENT, 12 * 5, NOTE_OFF, 12 * 5 + 7, NO_EVENT]
self.assertEqual(expected, list(melody))
# MonophonicMelody transposed up 19 half steps. 2 octave range.
events = [12 * 5 + 4, NO_EVENT, 12 * 5 + 5, NOTE_OFF, 12 * 6, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.transpose(transpose_amount=19, min_note=12 * 5, max_note=12 * 7)
expected = [12 * 6 + 11, NO_EVENT, 12 * 6, NOTE_OFF, 12 * 6 + 7, NO_EVENT]
self.assertEqual(expected, list(melody))
# MonophonicMelody transposed zero half steps. 1 octave range.
events = [12 * 4 + 11, 12 * 5, 12 * 5 + 11, NOTE_OFF, 12 * 6, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.transpose(transpose_amount=0, min_note=12 * 5, max_note=12 * 6)
expected = [12 * 5 + 11, 12 * 5, 12 * 5 + 11, NOTE_OFF, 12 * 5, NO_EVENT]
self.assertEqual(expected, list(melody))
def testSquash(self):
# MonophonicMelody in C, transposed to C, and squashed to 1 octave.
events = [12 * 5, NO_EVENT, 12 * 5 + 2, NOTE_OFF, 12 * 6 + 4, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 5, max_note=12 * 6, transpose_to_key=0)
expected = [12 * 5, NO_EVENT, 12 * 5 + 2, NOTE_OFF, 12 * 5 + 4, NO_EVENT]
self.assertEqual(expected, list(melody))
# MonophonicMelody in D, transposed to C, and squashed to 1 octave.
events = [12 * 5 + 2, 12 * 5 + 4, 12 * 6 + 7, 12 * 6 + 6, 12 * 5 + 1]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 5, max_note=12 * 6, transpose_to_key=0)
expected = [12 * 5, 12 * 5 + 2, 12 * 5 + 5, 12 * 5 + 4, 12 * 5 + 11]
self.assertEqual(expected, list(melody))
# MonophonicMelody in D, transposed to E, and squashed to 1 octave.
events = [12 * 5 + 2, 12 * 5 + 4, 12 * 6 + 7, 12 * 6 + 6, 12 * 4 + 11]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 5, max_note=12 * 6, transpose_to_key=4)
expected = [12 * 5 + 4, 12 * 5 + 6, 12 * 5 + 9, 12 * 5 + 8, 12 * 5 + 1]
self.assertEqual(expected, list(melody))
def testSquashCenterOctaves(self):
# Move up an octave.
events = [12 * 4, NO_EVENT, 12 * 4 + 2, NOTE_OFF, 12 * 4 + 4, NO_EVENT,
12 * 4 + 5, 12 * 5 + 2, 12 * 4 - 1, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 4, max_note=12 * 7, transpose_to_key=0)
expected = [12 * 5, NO_EVENT, 12 * 5 + 2, NOTE_OFF, 12 * 5 + 4, NO_EVENT,
12 * 5 + 5, 12 * 6 + 2, 12 * 5 - 1, NOTE_OFF]
self.assertEqual(expected, list(melody))
# Move down an octave.
events = [12 * 6, NO_EVENT, 12 * 6 + 2, NOTE_OFF, 12 * 6 + 4, NO_EVENT,
12 * 6 + 5, 12 * 7 + 2, 12 * 6 - 1, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 4, max_note=12 * 7, transpose_to_key=0)
expected = [12 * 5, NO_EVENT, 12 * 5 + 2, NOTE_OFF, 12 * 5 + 4, NO_EVENT,
12 * 5 + 5, 12 * 6 + 2, 12 * 5 - 1, NOTE_OFF]
self.assertEqual(expected, list(melody))
def testSquashMaxNote(self):
events = [12 * 5, 12 * 5 + 2, 12 * 5 + 4, 12 * 5 + 5, 12 * 5 + 11, 12 * 6,
12 * 6 + 1]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 5, max_note=12 * 6, transpose_to_key=0)
expected = [12 * 5, 12 * 5 + 2, 12 * 5 + 4, 12 * 5 + 5, 12 * 5 + 11, 12 * 5,
12 * 5 + 1]
self.assertEqual(expected, list(melody))
def testSquashAllNotesOff(self):
events = [NO_EVENT, NOTE_OFF, NO_EVENT, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.squash(min_note=12 * 4, max_note=12 * 7, transpose_to_key=0)
self.assertEqual(events, list(melody))
def testFromQuantizedSequence(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 0, 40), (11, 55, 1, 2), (40, 45, 10, 14),
(55, 120, 16, 17), (52, 99, 19, 20)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0)
expected = ([12, 11, NOTE_OFF, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT,
NO_EVENT, NO_EVENT, NO_EVENT, 40, NO_EVENT, NO_EVENT, NO_EVENT,
NOTE_OFF, NO_EVENT, 55, NOTE_OFF, NO_EVENT, 52])
self.assertEqual(expected, list(melody))
self.assertEqual(16, melody.steps_per_bar)
def testFromQuantizedSequenceNotCommonTimeSig(self):
self.quantized_sequence.time_signature = sequences_lib.TimeSignature(7, 8)
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 0, 40), (11, 55, 1, 2), (40, 45, 10, 14),
(55, 120, 16, 17), (52, 99, 19, 20)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0)
expected = ([12, 11, NOTE_OFF, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT,
NO_EVENT, NO_EVENT, NO_EVENT, 40, NO_EVENT, NO_EVENT, NO_EVENT,
NOTE_OFF, NO_EVENT, 55, NOTE_OFF, NO_EVENT, 52])
self.assertEqual(expected, list(melody))
self.assertEqual(14, melody.steps_per_bar)
def testFromNotesPolyphonic(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 4, 16), (19, 100, 4, 12)])
melody = melodies_lib.MonophonicMelody()
with self.assertRaises(melodies_lib.PolyphonicMelodyException):
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=False)
self.assertFalse(list(melody))
def testFromNotesPolyphonicWithIgnorePolyphonicNotes(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 0, 8), (19, 100, 0, 12),
(12, 100, 4, 12), (19, 100, 4, 16)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=True)
expected = ([19] + [NO_EVENT] * 3 + [19] + [NO_EVENT] * 11)
self.assertEqual(expected, list(melody))
self.assertEqual(16, melody.steps_per_bar)
def testFromNotesChord(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 4, 5), (19, 100, 4, 5),
(20, 100, 4, 5), (25, 100, 4, 5)])
melody = melodies_lib.MonophonicMelody()
with self.assertRaises(melodies_lib.PolyphonicMelodyException):
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=False)
self.assertFalse(list(melody))
def testFromNotesTrimEmptyMeasures(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 6, 7), (11, 100, 8, 9)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=False)
expected = [NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, 12,
NOTE_OFF, 11]
self.assertEqual(expected, list(melody))
self.assertEqual(16, melody.steps_per_bar)
def testFromNotesTimeOverlap(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 4, 8), (11, 100, 13, 15),
(13, 100, 8, 16)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=False)
expected = [NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, 12, NO_EVENT, NO_EVENT,
NO_EVENT, 13, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, 11,
NO_EVENT]
self.assertEqual(expected, list(melody))
def testFromNotesStepsPerBar(self):
self.quantized_sequence.time_signature = sequences_lib.TimeSignature(7, 8)
self.quantized_sequence.steps_per_quarter = 12
self.quantized_sequence.tracks[0] = []
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=0, track=0,
ignore_polyphonic_notes=False)
self.assertEqual(42, melody.steps_per_bar)
def testFromNotesStartAndEndStep(self):
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 4, 8), (11, 100, 9, 10), (13, 100, 13, 15),
(14, 100, 19, 20), (15, 100, 21, 27)])
melody = melodies_lib.MonophonicMelody()
melody.from_quantized_sequence(self.quantized_sequence,
start_step=18, track=0,
ignore_polyphonic_notes=False)
expected = [NO_EVENT, NO_EVENT, NO_EVENT, 14, NOTE_OFF, 15, NO_EVENT,
NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT]
self.assertEqual(expected, list(melody))
self.assertEqual(16, melody.start_step)
self.assertEqual(27, melody.end_step)
def testSetLength(self):
events = [60]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events, start_step=9)
melody.set_length(5)
self.assertListEqual([60, NOTE_OFF, NO_EVENT, NO_EVENT, NO_EVENT],
list(melody))
self.assertEquals(9, melody.start_step)
self.assertEquals(14, melody.end_step)
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events, start_step=9)
melody.set_length(5, from_left=True)
self.assertListEqual([NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, 60],
list(melody))
self.assertEquals(5, melody.start_step)
self.assertEquals(10, melody.end_step)
events = [60, NO_EVENT, NO_EVENT, NOTE_OFF]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.set_length(3)
self.assertListEqual([60, NO_EVENT, NO_EVENT], list(melody))
self.assertEquals(0, melody.start_step)
self.assertEquals(3, melody.end_step)
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
melody.set_length(3, from_left=True)
self.assertListEqual([NO_EVENT, NO_EVENT, NOTE_OFF], list(melody))
self.assertEquals(1, melody.start_step)
self.assertEquals(4, melody.end_step)
def testToSequenceSimple(self):
melody = melodies_lib.MonophonicMelody()
melody.from_event_list([NO_EVENT, 1, NO_EVENT, NOTE_OFF, NO_EVENT, 2, 3,
NOTE_OFF, NO_EVENT])
sequence = melody.to_sequence(
velocity=10,
instrument=1,
sequence_start_time=2,
qpm=60.0)
self.assertProtoEquals(
'ticks_per_quarter: 96 '
'tempos < qpm: 60.0 > '
'total_time: 3.75 '
'notes < '
' pitch: 1 velocity: 10 instrument: 1 start_time: 2.25 end_time: 2.75 '
'> '
'notes < '
' pitch: 2 velocity: 10 instrument: 1 start_time: 3.25 end_time: 3.5 '
'> '
'notes < '
' pitch: 3 velocity: 10 instrument: 1 start_time: 3.5 end_time: 3.75 '
'> ',
sequence)
def testToSequenceEndsWithSustainedNote(self):
melody = melodies_lib.MonophonicMelody()
melody.from_event_list([NO_EVENT, 1, NO_EVENT, NOTE_OFF, NO_EVENT, 2, 3,
NO_EVENT, NO_EVENT])
sequence = melody.to_sequence(
velocity=100,
instrument=0,
sequence_start_time=0,
qpm=60.0)
self.assertProtoEquals(
'ticks_per_quarter: 96 '
'tempos < qpm: 60.0 > '
'total_time: 2.25 '
'notes < pitch: 1 velocity: 100 start_time: 0.25 end_time: 0.75 > '
'notes < pitch: 2 velocity: 100 start_time: 1.25 end_time: 1.5 > '
'notes < pitch: 3 velocity: 100 start_time: 1.5 end_time: 2.25 > ',
sequence)
def testToSequenceEndsWithNonzeroStart(self):
melody = melodies_lib.MonophonicMelody()
melody.from_event_list([NO_EVENT, 1, NO_EVENT], start_step=4)
sequence = melody.to_sequence(
velocity=100,
instrument=0,
sequence_start_time=0.5,
qpm=60.0)
self.assertProtoEquals(
'ticks_per_quarter: 96 '
'tempos < qpm: 60.0 > '
'total_time: 2.25 '
'notes < pitch: 1 velocity: 100 start_time: 1.75 end_time: 2.25 > ',
sequence)
def testToSequenceEmpty(self):
melody = melodies_lib.MonophonicMelody()
sequence = melody.to_sequence(
velocity=10,
instrument=1,
sequence_start_time=2,
qpm=60.0)
self.assertProtoEquals(
'ticks_per_quarter: 96 '
'tempos < qpm: 60.0 > ',
sequence)
def testExtractMelodiesSimple(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 2, 4), (11, 1, 6, 7)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 9)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 11],
[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True)
self.assertEqual(2, len(melodies))
self.assertTrue(isinstance(melodies[0], melodies_lib.MonophonicMelody))
self.assertTrue(isinstance(melodies[1], melodies_lib.MonophonicMelody))
melodies = sorted([list(melody) for melody in melodies])
self.assertEqual(expected, melodies)
def testExtractMultipleMelodiesFromSameTrack(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 2, 4), (11, 1, 6, 11)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 8),
(50, 100, 33, 37), (52, 100, 34, 37)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 11,
NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT],
[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT],
[NO_EVENT, 50, 52, NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=2, min_unique_pitches=2,
ignore_polyphonic_notes=True)
melodies = sorted([list(melody) for melody in melodies])
self.assertEqual(expected, melodies)
def testExtractMelodiesMelodyTooShort(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 127, 2, 4), (14, 50, 6, 7)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 9)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=2, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True)
melodies = [list(melody) for melody in melodies]
self.assertEqual(expected, melodies)
def testExtractMelodiesPadEnd(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 127, 2, 4), (14, 50, 6, 7)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 8)])
testing_lib.add_quantized_track(
self.quantized_sequence, 2,
[(12, 127, 2, 4), (14, 50, 6, 9)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NOTE_OFF],
[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT],
[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT, NO_EVENT, NOTE_OFF, NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True, pad_end=True)
melodies = [list(melody) for melody in melodies]
self.assertEqual(expected, melodies)
def testExtractMelodiesMelodyTooLong(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 127, 2, 4), (14, 50, 6, 15)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 18)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14] +
[NO_EVENT] * 7,
[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14] +
[NO_EVENT] * 7]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, max_steps_truncate=14,
max_steps_discard=18, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True)
melodies = [list(melody) for melody in melodies]
self.assertEqual(expected, melodies)
def testExtractMelodiesMelodyTooLongWithPad(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 127, 2, 4), (14, 50, 6, 15)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 6, 18)])
expected = [[NO_EVENT, NO_EVENT, 12, NO_EVENT, NOTE_OFF, NO_EVENT, 14,
NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, max_steps_truncate=14,
max_steps_discard=18, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True, pad_end=True)
melodies = [list(melody) for melody in melodies]
self.assertEqual(expected, melodies)
def testExtractMelodiesTooFewPitches(self):
# Test that extract_melodies discards melodies with too few pitches where
# pitches are equivalent by octave.
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 0, 1), (13, 100, 1, 2), (18, 100, 2, 3),
(24, 100, 3, 4), (25, 100, 4, 5)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 100, 0, 1), (13, 100, 1, 2), (18, 100, 2, 3),
(25, 100, 3, 4), (26, 100, 4, 5)])
expected = [[12, 13, 18, 25, 26]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=1, min_unique_pitches=4,
ignore_polyphonic_notes=True)
melodies = [list(melody) for melody in melodies]
self.assertEqual(expected, melodies)
def testExtractMelodiesLateStart(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 102, 103), (13, 100, 104, 106)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 100, 100, 101), (13, 100, 102, 105)])
expected = [[NO_EVENT, NO_EVENT, 12, NOTE_OFF, 13, NO_EVENT],
[12, NOTE_OFF, 13, NO_EVENT, NO_EVENT]]
melodies, _ = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=True)
melodies = sorted([list(melody) for melody in melodies])
self.assertEqual(expected, melodies)
def testExtractMelodiesStatistics(self):
self.quantized_sequence.steps_per_quarter = 1
testing_lib.add_quantized_track(
self.quantized_sequence, 0,
[(12, 100, 2, 4), (11, 1, 6, 7), (10, 100, 8, 10), (9, 100, 11, 14),
(8, 100, 16, 40), (7, 100, 41, 42)])
testing_lib.add_quantized_track(
self.quantized_sequence, 1,
[(12, 127, 2, 4), (14, 50, 2, 8)])
testing_lib.add_quantized_track(
self.quantized_sequence, 2,
[(12, 127, 0, 1)])
testing_lib.add_quantized_track(
self.quantized_sequence, 3,
[(12, 127, 2, 4), (12, 50, 6, 8)])
_, stats = melodies_lib.extract_melodies(
self.quantized_sequence, min_bars=1, gap_bars=1, min_unique_pitches=2,
ignore_polyphonic_notes=False)
stats_dict = dict([(stat.name, stat) for stat in stats])
self.assertEqual(stats_dict['polyphonic_tracks_discarded'].count, 1)
self.assertEqual(stats_dict['melodies_discarded_too_short'].count, 1)
self.assertEqual(stats_dict['melodies_discarded_too_few_pitches'].count, 1)
self.assertEqual(
stats_dict['melody_lengths_in_bars'].counters,
{float('-inf'): 0, 0: 1, 1: 0, 2: 1, 10: 1, 20: 0, 30: 0, 40: 0, 50: 0,
100: 0, 200: 0, 500: 0})
class MelodyEncoderDecoderTest(tf.test.TestCase):
def setUp(self):
self.melody_encoder_decoder = melodies_lib.OneHotEncoderDecoder(60, 72, 0)
def testMinNoteMaxNoteAndTransposeToKeyValidValues(self):
# Valid parameters
melodies_lib.OneHotEncoderDecoder(0, 128, 0)
melodies_lib.OneHotEncoderDecoder(60, 72, 11)
# Invalid parameters
self.assertRaises(ValueError, melodies_lib.OneHotEncoderDecoder, -1, 72, 0)
self.assertRaises(ValueError, melodies_lib.OneHotEncoderDecoder, 60, 129, 0)
self.assertRaises(ValueError, melodies_lib.OneHotEncoderDecoder, 60, 71, 0)
self.assertRaises(ValueError, melodies_lib.OneHotEncoderDecoder, 60, 72, -1)
self.assertRaises(ValueError, melodies_lib.OneHotEncoderDecoder, 60, 72, 12)
def testInitValues(self):
self.assertEqual(self.melody_encoder_decoder.min_note, 60)
self.assertEqual(self.melody_encoder_decoder.max_note, 72)
self.assertEqual(self.melody_encoder_decoder.transpose_to_key, 0)
self.assertEqual(self.melody_encoder_decoder.input_size, 14)
self.assertEqual(self.melody_encoder_decoder.num_classes, 14)
self.assertEqual(self.melody_encoder_decoder.no_event_label, 0)
def testSquashAndEncode(self):
events = [100, 100, 107, 111, NO_EVENT, 99, 112, NOTE_OFF, NO_EVENT]
melody = melodies_lib.MonophonicMelody()
melody.from_event_list(events)
sequence_example = self.melody_encoder_decoder.squash_and_encode(melody)
expected_inputs = [
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]]
expected_labels = [2, 9, 13, 0, 13, 2, 1, 0]
expected_sequence_example = sequence_example_lib.make_sequence_example(
expected_inputs, expected_labels)
self.assertEqual(sequence_example, expected_sequence_example)
def testGetInputsBatch(self):
events1 = [100, 100, 107, 111, NO_EVENT, 99, 112, NOTE_OFF, NO_EVENT]
melody1 = melodies_lib.MonophonicMelody()
melody1.from_event_list(events1)
events2 = [9, 10, 12, 14, 15, 17, 19, 21, 22]
melody2 = melodies_lib.MonophonicMelody()
melody2.from_event_list(events2)
transpose_amount1 = melody1.squash(
self.melody_encoder_decoder.min_note,
self.melody_encoder_decoder.max_note,
self.melody_encoder_decoder.transpose_to_key)
transpose_amount2 = melody2.squash(
self.melody_encoder_decoder.min_note,
self.melody_encoder_decoder.max_note,
self.melody_encoder_decoder.transpose_to_key)
self.assertEqual(transpose_amount1, -40)
self.assertEqual(transpose_amount2, 50)
melodies = [melody1, melody2]
expected_inputs1 = [
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]]
expected_inputs2 = [
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0],
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]]
expected_full_length_inputs_batch = [expected_inputs1, expected_inputs2]
expected_last_event_inputs_batch = [expected_inputs1[-1:],
expected_inputs2[-1:]]
self.assertListEqual(
expected_full_length_inputs_batch,
self.melody_encoder_decoder.get_inputs_batch(melodies, True))
self.assertListEqual(
expected_last_event_inputs_batch,
self.melody_encoder_decoder.get_inputs_batch(melodies))
def testExtendMelodies(self):
melody1 = melodies_lib.MonophonicMelody()
melody1.from_event_list([60])
melody2 = melodies_lib.MonophonicMelody()
melody2.from_event_list([60])
melody3 = melodies_lib.MonophonicMelody()
melody3.from_event_list([60])
melody4 = melodies_lib.MonophonicMelody()
melody4.from_event_list([60])
melodies = [melody1, melody2, melody3, melody4]
softmax = [[
[0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
], [
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0]
], [
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
], [
[0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
]]
self.melody_encoder_decoder.extend_event_sequences(melodies, softmax)
self.assertListEqual(list(melody1), [60, 60])
self.assertListEqual(list(melody2), [60, 71])
self.assertListEqual(list(melody3), [60, NO_EVENT])
self.assertListEqual(list(melody4), [60, NOTE_OFF])
if __name__ == '__main__':
tf.test.main()