mir-evaluation · craffel · Feb 13, 2017 · Feb 9, 2017 · Feb 9, 2017 · Feb 9, 2017
diff --git a/.travis.yml b/.travis.yml
@@ -27,13 +27,13 @@ before_install:
     - pip install numpydoc
 
 install:
-    - pip install -e .[display]
+    - pip install -e .[display,testing]
 
 before_script:
     - pep8 mir_eval evaluators tests
 
 script:
-    - nosetests -v --with-coverage --cover-package=mir_eval
+    - nosetests -v --with-coverage --cover-package=mir_eval -w tests
     - python setup.py build_sphinx
     - python setup.py egg_info -b.dev sdist --formats gztar
 

diff --git a/setup.py b/setup.py
@@ -31,6 +31,7 @@
     ],
     extras_require={
         'display': ['matplotlib>=1.5.0',
-                    'scipy>=0.16.0']
+                    'scipy>=0.16.0'],
+        'testing': ['matplotlib>=2.0.0']
     }
 )
diff --git a/tests/__init__.py b/tests/__init__.py
diff --git a/tests/baseline_images/test_display/events.png b/tests/baseline_images/test_display/events.png
diff --git a/tests/baseline_images/test_display/hierarchy_label.png b/tests/baseline_images/test_display/hierarchy_label.png
diff --git a/tests/baseline_images/test_display/hierarchy_nolabel.png b/tests/baseline_images/test_display/hierarchy_nolabel.png
diff --git a/tests/baseline_images/test_display/labeled_events.png b/tests/baseline_images/test_display/labeled_events.png
diff --git a/tests/baseline_images/test_display/labeled_intervals.png b/tests/baseline_images/test_display/labeled_intervals.png
diff --git a/tests/baseline_images/test_display/labeled_intervals_compare.png b/tests/baseline_images/test_display/labeled_intervals_compare.png
diff --git a/tests/baseline_images/test_display/labeled_intervals_compare_common.png b/tests/baseline_images/test_display/labeled_intervals_compare_common.png
diff --git a/tests/baseline_images/test_display/labeled_intervals_compare_noextend.png b/tests/baseline_images/test_display/labeled_intervals_compare_noextend.png
diff --git a/tests/baseline_images/test_display/labeled_intervals_noextend.png b/tests/baseline_images/test_display/labeled_intervals_noextend.png
diff --git a/tests/baseline_images/test_display/multipitch_hz_unvoiced.png b/tests/baseline_images/test_display/multipitch_hz_unvoiced.png
diff --git a/tests/baseline_images/test_display/multipitch_hz_voiced.png b/tests/baseline_images/test_display/multipitch_hz_voiced.png
diff --git a/tests/baseline_images/test_display/multipitch_midi.png b/tests/baseline_images/test_display/multipitch_midi.png
diff --git a/tests/baseline_images/test_display/piano_roll.png b/tests/baseline_images/test_display/piano_roll.png
diff --git a/tests/baseline_images/test_display/piano_roll_midi.png b/tests/baseline_images/test_display/piano_roll_midi.png
diff --git a/tests/baseline_images/test_display/pitch_hz.png b/tests/baseline_images/test_display/pitch_hz.png
diff --git a/tests/baseline_images/test_display/pitch_midi.png b/tests/baseline_images/test_display/pitch_midi.png
diff --git a/tests/baseline_images/test_display/pitch_midi_hz.png b/tests/baseline_images/test_display/pitch_midi_hz.png
diff --git a/tests/baseline_images/test_display/segment.png b/tests/baseline_images/test_display/segment.png
diff --git a/tests/baseline_images/test_display/segment_text.png b/tests/baseline_images/test_display/segment_text.png
diff --git a/tests/baseline_images/test_display/separation.png b/tests/baseline_images/test_display/separation.png
diff --git a/tests/baseline_images/test_display/separation_label.png b/tests/baseline_images/test_display/separation_label.png
diff --git a/tests/baseline_images/test_display/ticker_midi_zoom.png b/tests/baseline_images/test_display/ticker_midi_zoom.png
diff --git a/tests/baseline_images/tests.test_display/events.png b/tests/baseline_images/tests.test_display/events.png
diff --git a/tests/baseline_images/tests.test_display/hierarchy_label.png b/tests/baseline_images/tests.test_display/hierarchy_label.png
diff --git a/tests/baseline_images/tests.test_display/hierarchy_nolabel.png b/tests/baseline_images/tests.test_display/hierarchy_nolabel.png
diff --git a/tests/baseline_images/tests.test_display/labeled_events.png b/tests/baseline_images/tests.test_display/labeled_events.png
diff --git a/tests/baseline_images/tests.test_display/labeled_intervals.png b/tests/baseline_images/tests.test_display/labeled_intervals.png
diff --git a/tests/baseline_images/tests.test_display/labeled_intervals_compare.png b/tests/baseline_images/tests.test_display/labeled_intervals_compare.png
diff --git a/tests/baseline_images/tests.test_display/labeled_intervals_compare_common.png b/tests/baseline_images/tests.test_display/labeled_intervals_compare_common.png
diff --git a/tests/baseline_images/tests.test_display/labeled_intervals_compare_noextend.png b/tests/baseline_images/tests.test_display/labeled_intervals_compare_noextend.png
diff --git a/tests/baseline_images/tests.test_display/labeled_intervals_noextend.png b/tests/baseline_images/tests.test_display/labeled_intervals_noextend.png
diff --git a/tests/baseline_images/tests.test_display/multipitch_hz_unvoiced.png b/tests/baseline_images/tests.test_display/multipitch_hz_unvoiced.png
diff --git a/tests/baseline_images/tests.test_display/multipitch_hz_voiced.png b/tests/baseline_images/tests.test_display/multipitch_hz_voiced.png
diff --git a/tests/baseline_images/tests.test_display/multipitch_midi.png b/tests/baseline_images/tests.test_display/multipitch_midi.png
diff --git a/tests/baseline_images/tests.test_display/piano_roll.png b/tests/baseline_images/tests.test_display/piano_roll.png
diff --git a/tests/baseline_images/tests.test_display/piano_roll_midi.png b/tests/baseline_images/tests.test_display/piano_roll_midi.png
diff --git a/tests/baseline_images/tests.test_display/pitch_hz.png b/tests/baseline_images/tests.test_display/pitch_hz.png
diff --git a/tests/baseline_images/tests.test_display/pitch_midi.png b/tests/baseline_images/tests.test_display/pitch_midi.png
diff --git a/tests/baseline_images/tests.test_display/pitch_midi_hz.png b/tests/baseline_images/tests.test_display/pitch_midi_hz.png
diff --git a/tests/baseline_images/tests.test_display/segment.png b/tests/baseline_images/tests.test_display/segment.png
diff --git a/tests/baseline_images/tests.test_display/segment_text.png b/tests/baseline_images/tests.test_display/segment_text.png
diff --git a/tests/baseline_images/tests.test_display/separation.png b/tests/baseline_images/tests.test_display/separation.png
diff --git a/tests/baseline_images/tests.test_display/separation_label.png b/tests/baseline_images/tests.test_display/separation_label.png
diff --git a/tests/baseline_images/tests.test_display/ticker_midi_zoom.png b/tests/baseline_images/tests.test_display/ticker_midi_zoom.png
diff --git a/tests/generate_data.py b/tests/generate_data.py
@@ -52,27 +52,27 @@ def load_separation_data(folder):
     # that task will be generated.
     tasks = {}
     tasks['beat'] = (mir_eval.beat, mir_eval.io.load_events,
-                     'tests/data/beat/{}*.txt')
+                     'data/beat/{}*.txt')
     tasks['chord'] = (mir_eval.chord, mir_eval.io.load_labeled_intervals,
-                      'tests/data/chord/{}*.lab')
+                      'data/chord/{}*.lab')
     tasks['melody'] = (mir_eval.melody, mir_eval.io.load_time_series,
-                       'tests/data/melody/{}*.txt')
+                       'data/melody/{}*.txt')
     tasks['multipitch'] = (mir_eval.multipitch,
                            mir_eval.io.load_ragged_time_series,
-                           'tests/data/multipitch/()*.txt')
+                           'data/multipitch/()*.txt')
     tasks['onset'] = (mir_eval.onset, mir_eval.io.load_events,
-                      'tests/data/onset/{}*.txt')
+                      'data/onset/{}*.txt')
     tasks['pattern'] = (mir_eval.pattern, mir_eval.io.load_patterns,
-                        'tests/data/pattern/{}*.txt')
+                        'data/pattern/{}*.txt')
     tasks['segment'] = (mir_eval.segment, mir_eval.io.load_labeled_intervals,
-                        'tests/data/segment/{}*.lab')
+                        'data/segment/{}*.lab')
     tasks['separation'] = (mir_eval.separation, load_separation_data,
-                           'tests/data/separation/{}*')
+                           'data/separation/{}*')
     tasks['transcription'] = (mir_eval.transcription,
                               mir_eval.io.load_valued_intervals,
-                              'tests/data/transcription/{}*.txt')
+                              'data/transcription/{}*.txt')
     tasks['key'] = (mir_eval.key, mir_eval.io.load_key,
-                    'tests/data/key/{}*.txt')
+                    'data/key/{}*.txt')
     # Get task keys from argv
     for task in sys.argv[1:]:
         print('Generating data for {}'.format(task))

diff --git a/tests/test_beat.py b/tests/test_beat.py
@@ -12,9 +12,9 @@
 A_TOL = 1e-12
 
 # Path to the fixture files
-REF_GLOB = 'tests/data/beat/ref*.txt'
-EST_GLOB = 'tests/data/beat/est*.txt'
-SCORES_GLOB = 'tests/data/beat/output*.json'
+REF_GLOB = 'data/beat/ref*.txt'
+EST_GLOB = 'data/beat/est*.txt'
+SCORES_GLOB = 'data/beat/output*.json'
 
 
 def test_trim_beats():

diff --git a/tests/test_chord.py b/tests/test_chord.py
@@ -12,9 +12,9 @@
 A_TOL = 1e-12
 
 # Path to the fixture files
-REF_GLOB = 'tests/data/chord/ref*.lab'
-EST_GLOB = 'tests/data/chord/est*.lab'
-SCORES_GLOB = 'tests/data/chord/output*.json'
+REF_GLOB = 'data/chord/ref*.lab'
+EST_GLOB = 'data/chord/est*.lab'
+SCORES_GLOB = 'data/chord/output*.json'
 
 
 def __check_valid(function, parameters, result):

diff --git a/tests/test_display.py b/tests/test_display.py
@@ -8,11 +8,10 @@
 matplotlib.use('Agg')  # nopep8
 
 import matplotlib.pyplot as plt
-from matplotlib.style import context as style_context
 import numpy as np
 
 # Import the hacked image comparison module
-from tests.mpl_ic import image_comparison
+from mpl_ic import image_comparison
 
 from nose.tools import raises
 
@@ -28,15 +27,10 @@
 from mir_eval.io import load_wav
 
 
-# Test fixtures will use seaborn-muted style
-GLOBAL_STYLE = 'seaborn-muted'
-
-
 @decorator
 def styled(f, *args, **kwargs):
-    matplotlib.rcParams.update(matplotlib.rcParamsDefault)
-    with style_context(GLOBAL_STYLE):
-        return f(*args, **kwargs)
+    matplotlib.rcdefaults()
+    return f(*args, **kwargs)
 
 
 @image_comparison(baseline_images=['segment'], extensions=['png'])
@@ -46,7 +40,7 @@ def test_display_segment():
     plt.figure()
 
     # Load some segment data
-    intervals, labels = load_labeled_intervals('tests/data/segment/ref00.lab')
+    intervals, labels = load_labeled_intervals('data/segment/ref00.lab')
 
     # Plot the segments with no labels
     mir_eval.display.segments(intervals, labels, text=False)
@@ -61,7 +55,7 @@ def test_display_segment_text():
     plt.figure()
 
     # Load some segment data
-    intervals, labels = load_labeled_intervals('tests/data/segment/ref00.lab')
+    intervals, labels = load_labeled_intervals('data/segment/ref00.lab')
 
     # Plot the segments with no labels
     mir_eval.display.segments(intervals, labels, text=True)
@@ -74,7 +68,7 @@ def test_display_labeled_intervals():
     plt.figure()
 
     # Load some chord data
-    intervals, labels = load_labeled_intervals('tests/data/chord/ref01.lab')
+    intervals, labels = load_labeled_intervals('data/chord/ref01.lab')
 
     # Plot the chords with nothing fancy
     mir_eval.display.labeled_intervals(intervals, labels)
@@ -88,7 +82,7 @@ def test_display_labeled_intervals_noextend():
     plt.figure()
 
     # Load some chord data
-    intervals, labels = load_labeled_intervals('tests/data/chord/ref01.lab')
+    intervals, labels = load_labeled_intervals('data/chord/ref01.lab')
 
     # Plot the chords with nothing fancy
     ax = plt.axes()
@@ -107,8 +101,8 @@ def test_display_labeled_intervals_compare():
     plt.figure()
 
     # Load some chord data
-    ref_int, ref_labels = load_labeled_intervals('tests/data/chord/ref01.lab')
-    est_int, est_labels = load_labeled_intervals('tests/data/chord/est01.lab')
+    ref_int, ref_labels = load_labeled_intervals('data/chord/ref01.lab')
+    est_int, est_labels = load_labeled_intervals('data/chord/est01.lab')
 
     # Plot reference and estimates using label set extension
     mir_eval.display.labeled_intervals(ref_int, ref_labels,
@@ -127,8 +121,8 @@ def test_display_labeled_intervals_compare_noextend():
     plt.figure()
 
     # Load some chord data
-    ref_int, ref_labels = load_labeled_intervals('tests/data/chord/ref01.lab')
-    est_int, est_labels = load_labeled_intervals('tests/data/chord/est01.lab')
+    ref_int, ref_labels = load_labeled_intervals('data/chord/ref01.lab')
+    est_int, est_labels = load_labeled_intervals('data/chord/est01.lab')
 
     # Plot reference and estimate, but only use the reference labels
     mir_eval.display.labeled_intervals(ref_int, ref_labels,
@@ -148,8 +142,8 @@ def test_display_labeled_intervals_compare_common():
     plt.figure()
 
     # Load some chord data
-    ref_int, ref_labels = load_labeled_intervals('tests/data/chord/ref01.lab')
-    est_int, est_labels = load_labeled_intervals('tests/data/chord/est01.lab')
+    ref_int, ref_labels = load_labeled_intervals('data/chord/ref01.lab')
+    est_int, est_labels = load_labeled_intervals('data/chord/est01.lab')
 
     label_set = list(sorted(set(ref_labels) | set(est_labels)))
 
@@ -171,8 +165,8 @@ def test_display_hierarchy_nolabel():
     plt.figure()
 
     # Load some chord data
-    int0, lab0 = load_labeled_intervals('tests/data/hierarchy/ref00.lab')
-    int1, lab1 = load_labeled_intervals('tests/data/hierarchy/ref01.lab')
+    int0, lab0 = load_labeled_intervals('data/hierarchy/ref00.lab')
+    int1, lab1 = load_labeled_intervals('data/hierarchy/ref01.lab')
 
     # Plot reference and estimate with a common label set
     mir_eval.display.hierarchy([int0, int1],
@@ -188,8 +182,8 @@ def test_display_hierarchy_label():
     plt.figure()
 
     # Load some chord data
-    int0, lab0 = load_labeled_intervals('tests/data/hierarchy/ref00.lab')
-    int1, lab1 = load_labeled_intervals('tests/data/hierarchy/ref01.lab')
+    int0, lab0 = load_labeled_intervals('data/hierarchy/ref00.lab')
+    int1, lab1 = load_labeled_intervals('data/hierarchy/ref01.lab')
 
     # Plot reference and estimate with a common label set
     mir_eval.display.hierarchy([int0, int1],
@@ -204,8 +198,8 @@ def test_display_hierarchy_label():
 def test_pitch_hz():
     plt.figure()
 
-    ref_times, ref_freqs = load_labeled_events('tests/data/melody/ref00.txt')
-    est_times, est_freqs = load_labeled_events('tests/data/melody/est00.txt')
+    ref_times, ref_freqs = load_labeled_events('data/melody/ref00.txt')
+    est_times, est_freqs = load_labeled_events('data/melody/est00.txt')
 
     # Plot pitches on a Hz scale
     mir_eval.display.pitch(ref_times, ref_freqs, unvoiced=True,
@@ -220,7 +214,7 @@ def test_pitch_hz():
 def test_pitch_midi():
     plt.figure()
 
-    times, freqs = load_labeled_events('tests/data/melody/ref00.txt')
+    times, freqs = load_labeled_events('data/melody/ref00.txt')
 
     # Plot pitches on a midi scale with note tickers
     mir_eval.display.pitch(times, freqs, midi=True)
@@ -232,7 +226,7 @@ def test_pitch_midi():
 def test_pitch_midi_hz():
     plt.figure()
 
-    times, freqs = load_labeled_events('tests/data/melody/ref00.txt')
+    times, freqs = load_labeled_events('data/melody/ref00.txt')
 
     # Plot pitches on a midi scale with note tickers
     mir_eval.display.pitch(times, freqs, midi=True)
@@ -245,7 +239,7 @@ def test_pitch_midi_hz():
 def test_multipitch_hz_unvoiced():
     plt.figure()
 
-    times, pitches = load_ragged_time_series('tests/data/multipitch/est01.txt')
+    times, pitches = load_ragged_time_series('data/multipitch/est01.txt')
 
     # Plot pitches on a midi scale with note tickers
     mir_eval.display.multipitch(times, pitches, midi=False, unvoiced=True)
@@ -256,7 +250,7 @@ def test_multipitch_hz_unvoiced():
 def test_multipitch_hz_voiced():
     plt.figure()
 
-    times, pitches = load_ragged_time_series('tests/data/multipitch/est01.txt')
+    times, pitches = load_ragged_time_series('data/multipitch/est01.txt')
 
     mir_eval.display.multipitch(times, pitches, midi=False, unvoiced=False)
 
@@ -266,8 +260,8 @@ def test_multipitch_hz_voiced():
 def test_multipitch_midi():
     plt.figure()
 
-    ref_t, ref_p = load_ragged_time_series('tests/data/multipitch/ref01.txt')
-    est_t, est_p = load_ragged_time_series('tests/data/multipitch/est01.txt')
+    ref_t, ref_p = load_ragged_time_series('data/multipitch/ref01.txt')
+    est_t, est_p = load_ragged_time_series('data/multipitch/est01.txt')
 
     # Plot pitches on a midi scale with note tickers
     mir_eval.display.multipitch(ref_t, ref_p, midi=True,
@@ -283,8 +277,8 @@ def test_multipitch_midi():
 def test_pianoroll():
     plt.figure()
 
-    ref_t, ref_p = load_valued_intervals('tests/data/transcription/ref04.txt')
-    est_t, est_p = load_valued_intervals('tests/data/transcription/est04.txt')
+    ref_t, ref_p = load_valued_intervals('data/transcription/ref04.txt')
+    est_t, est_p = load_valued_intervals('data/transcription/est04.txt')
 
     mir_eval.display.piano_roll(ref_t, ref_p,
                                 label='Reference', alpha=0.5)
@@ -299,8 +293,8 @@ def test_pianoroll():
 def test_pianoroll_midi():
     plt.figure()
 
-    ref_t, ref_p = load_valued_intervals('tests/data/transcription/ref04.txt')
-    est_t, est_p = load_valued_intervals('tests/data/transcription/est04.txt')
+    ref_t, ref_p = load_valued_intervals('data/transcription/ref04.txt')
+    est_t, est_p = load_valued_intervals('data/transcription/est04.txt')
 
     ref_midi = mir_eval.util.hz_to_midi(ref_p)
     est_midi = mir_eval.util.hz_to_midi(est_p)
@@ -327,9 +321,9 @@ def test_ticker_midi_zoom():
 def test_separation():
     plt.figure()
 
-    x0, fs = load_wav('tests/data/separation/ref05/0.wav')
-    x1, fs = load_wav('tests/data/separation/ref05/1.wav')
-    x2, fs = load_wav('tests/data/separation/ref05/2.wav')
+    x0, fs = load_wav('data/separation/ref05/0.wav')
+    x1, fs = load_wav('data/separation/ref05/1.wav')
+    x2, fs = load_wav('data/separation/ref05/2.wav')
 
     mir_eval.display.separation([x0, x1, x2], fs=fs)
 
@@ -339,9 +333,9 @@ def test_separation():
 def test_separation_label():
     plt.figure()
 
-    x0, fs = load_wav('tests/data/separation/ref05/0.wav')
-    x1, fs = load_wav('tests/data/separation/ref05/1.wav')
-    x2, fs = load_wav('tests/data/separation/ref05/2.wav')
+    x0, fs = load_wav('data/separation/ref05/0.wav')
+    x1, fs = load_wav('data/separation/ref05/1.wav')
+    x2, fs = load_wav('data/separation/ref05/2.wav')
 
     mir_eval.display.separation([x0, x1, x2], fs=fs,
                                 labels=['Alice', 'Bob', 'Carol'])
@@ -355,8 +349,8 @@ def test_events():
     plt.figure()
 
     # Load some event data
-    beats_ref = mir_eval.io.load_events('tests/data/beat/ref00.txt')[:30]
-    beats_est = mir_eval.io.load_events('tests/data/beat/est00.txt')[:30]
+    beats_ref = mir_eval.io.load_events('data/beat/ref00.txt')[:30]
+    beats_est = mir_eval.io.load_events('data/beat/est00.txt')[:30]
 
     # Plot both with labels
     mir_eval.display.events(beats_ref, label='reference')
@@ -370,7 +364,7 @@ def test_labeled_events():
     plt.figure()
 
     # Load some event data
-    beats_ref = mir_eval.io.load_events('tests/data/beat/ref00.txt')[:10]
+    beats_ref = mir_eval.io.load_events('data/beat/ref00.txt')[:10]
 
     labels = list('abcdefghijklmnop')
     # Plot both with labels

diff --git a/tests/test_hierarchy.py b/tests/test_hierarchy.py
@@ -118,9 +118,9 @@ def __test(window, frame_size):
 
 def test_tmeasure_regression():
 
-    ref_files = sorted(glob('tests/data/hierarchy/ref*.lab'))
-    est_files = sorted(glob('tests/data/hierarchy/est*.lab'))
-    out_files = sorted(glob('tests/data/hierarchy/output*.json'))
+    ref_files = sorted(glob('data/hierarchy/ref*.lab'))
+    est_files = sorted(glob('data/hierarchy/est*.lab'))
+    out_files = sorted(glob('data/hierarchy/output*.json'))
 
     ref_hier = [mir_eval.io.load_labeled_intervals(_) for _ in ref_files]
     est_hier = [mir_eval.io.load_labeled_intervals(_) for _ in est_files]

diff --git a/tests/test_input_output.py b/tests/test_input_output.py
@@ -146,27 +146,27 @@ def test_load_ragged_time_series():
 
 def test_load_tempo():
     # Test the tempo loader
-    tempi, weight = mir_eval.io.load_tempo('tests/data/tempo/ref01.lab')
+    tempi, weight = mir_eval.io.load_tempo('data/tempo/ref01.lab')
 
     assert np.allclose(tempi, [60, 120])
     assert weight == 0.5
 
 
 @nose.tools.raises(ValueError)
 def test_load_tempo_multiline():
-    tempi, weight = mir_eval.io.load_tempo('tests/data/tempo/bad00.lab')
+    tempi, weight = mir_eval.io.load_tempo('data/tempo/bad00.lab')
 
 
 @nose.tools.raises(ValueError)
 def test_load_tempo_badweight():
-    tempi, weight = mir_eval.io.load_tempo('tests/data/tempo/bad01.lab')
+    tempi, weight = mir_eval.io.load_tempo('data/tempo/bad01.lab')
 
 
 def test_load_bad_tempi():
 
     with warnings.catch_warnings(record=True) as w:
         warnings.simplefilter('always')
-        tempi, weight = mir_eval.io.load_tempo('tests/data/tempo/bad02.lab')
+        tempi, weight = mir_eval.io.load_tempo('data/tempo/bad02.lab')
 
         assert len(w) == 1
         assert issubclass(w[-1].category, UserWarning)

diff --git a/tests/test_key.py b/tests/test_key.py
@@ -11,9 +11,9 @@
 A_TOL = 1e-12
 
 # Path to the fixture files
-REF_GLOB = 'tests/data/key/ref*.txt'
-EST_GLOB = 'tests/data/key/est*.txt'
-SCORES_GLOB = 'tests/data/key/output*.json'
+REF_GLOB = 'data/key/ref*.txt'
+EST_GLOB = 'data/key/est*.txt'
+SCORES_GLOB = 'data/key/output*.json'
 
 
 def __unit_test_key_function(metric):