-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathtest_knn.py
287 lines (248 loc) · 9.93 KB
/
test_knn.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
from math import nan
from unittest import TestCase
import numpy as np
from scipy.io import wavfile
from sklearn.neighbors import KNeighborsRegressor
from entrainment_metrics import InterPausalUnit
from entrainment_metrics.continuous import TimeSeries, calculate_metric
class KNNTestCase(TestCase):
def setUp(self):
self.cases = {
'long_100-200-300_x2': {
'audio': wavfile.read("./data/100-200-300_long_x2.wav"),
'ipus': [
InterPausalUnit(0.0, 4.0, {'F0_MAX': 100.003}),
InterPausalUnit(8.0, 12.0, {'F0_MAX': 200.002}),
InterPausalUnit(16.0, 24.0, {'F0_MAX': 300.002}),
InterPausalUnit(28.0, 32.0, {'F0_MAX': 100.003}),
InterPausalUnit(36.0, 40.0, {'F0_MAX': 200.002}),
InterPausalUnit(44.0, 52.0, {'F0_MAX': 300.002}),
],
'ipus_middle_points_in_time': [2.0, 10.0, 20.0, 30.0, 38.0, 48.0],
'ipus_feature_values': [
100.003,
200.002,
300.002,
100.003,
200.002,
300.002,
],
},
'long_300-200-100_x2': {
'ipus': [
InterPausalUnit(0.0, 8.0, {'F0_MAX': 300.002}),
InterPausalUnit(12.0, 16.0, {'F0_MAX': 200.002}),
InterPausalUnit(20.0, 24.0, {'F0_MAX': 100.003}),
InterPausalUnit(28.0, 36.0, {'F0_MAX': 300.002}),
InterPausalUnit(40.0, 44.0, {'F0_MAX': 200.002}),
InterPausalUnit(48.0, 52.0, {'F0_MAX': 100.003}),
],
},
'ipu_with_None_value': {
'ipus': [
InterPausalUnit(0.0, 8.0, {'F0_MAX': None}),
InterPausalUnit(12.0, 16.0, {'F0_MAX': 200.002}),
InterPausalUnit(20.0, 24.0, {'F0_MAX': 100.003}),
InterPausalUnit(28.0, 36.0, {'F0_MAX': 300.002}),
InterPausalUnit(40.0, 44.0, {'F0_MAX': 200.002}),
InterPausalUnit(48.0, 52.0, {'F0_MAX': 100.003}),
],
},
'ipu_with_nan_value': {
'ipus': [
InterPausalUnit(0.0, 8.0, {'F0_MAX': 300.002}),
InterPausalUnit(12.0, 16.0, {'F0_MAX': 200.002}),
InterPausalUnit(20.0, 24.0, {'F0_MAX': 100.003}),
InterPausalUnit(28.0, 36.0, {'F0_MAX': nan}),
InterPausalUnit(40.0, 44.0, {'F0_MAX': 200.002}),
InterPausalUnit(48.0, 52.0, {'F0_MAX': 100.003}),
],
},
'unordered': {
'ipus': [
InterPausalUnit(16.0, 24.0, {'F0_MAX': 300.002}),
InterPausalUnit(0.0, 4.0, {'F0_MAX': 100.003}),
InterPausalUnit(8.0, 12.0, {'F0_MAX': 200.002}),
],
},
}
def test_calculate_knn_time_series_longx2(self):
case = self.cases['long_100-200-300_x2']
model = KNeighborsRegressor(n_neighbors=4)
X = np.array(case['ipus_middle_points_in_time'])
X = X.reshape(-1, 1)
y = case['ipus_feature_values']
model.fit(X, y)
samplerate, data = case['audio']
values_to_predict = [
i / samplerate
for i in range(2 * samplerate, 48 * samplerate, int(0.01 * samplerate))
]
time_series = TimeSeries(
feature='F0_MAX', interpausal_units=case['ipus'], method='knn', k=4
)
np.testing.assert_almost_equal(
model.predict(np.array(values_to_predict).reshape(-1, 1)),
time_series.predict(values_to_predict),
)
def test_calculate_knn_time_series_warnings_longx2(self):
case = self.cases['long_100-200-300_x2']
time_series = TimeSeries(
feature='F0_MAX', interpausal_units=case['ipus'], method='knn', k=4
)
values_before_start_to_predict = [-1.0]
values_after_end_to_predict = [53.0]
self.assertWarns(Warning, time_series.predict, values_before_start_to_predict)
self.assertWarns(Warning, time_series.predict, values_after_end_to_predict)
def test_calculate_proximity_with_itself(self):
case = self.cases['long_100-200-300_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case['ipus'], method='knn', k=4
)
self.assertEqual(
calculate_metric("proximity", time_series_a, time_series_a), 0.0
)
def test_calculate_convergence_with_itself(self):
case = self.cases['long_100-200-300_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case['ipus'], method='knn', k=4
)
self.assertTrue(
np.isnan(
calculate_metric("convergence", time_series_a, time_series_a),
)
)
def test_calculate_synchrony_with_itself(self):
case = self.cases['long_100-200-300_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case['ipus'], method='knn', k=4
)
self.assertEqual(
calculate_metric("synchrony", time_series_a, time_series_a), 1.0
)
def test_calculate_proximity_oposites(self):
case_a = self.cases['long_100-200-300_x2']
case_b = self.cases['long_300-200-100_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
time_series_b = TimeSeries(
feature='F0_MAX', interpausal_units=case_b['ipus'], method='knn', k=4
)
np.testing.assert_almost_equal(
calculate_metric("proximity", time_series_a, time_series_b),
-0.011360997500560188,
)
def test_calculate_convergence_oposites(self):
case_a = self.cases['long_100-200-300_x2']
case_b = self.cases['long_300-200-100_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
time_series_b = TimeSeries(
feature='F0_MAX', interpausal_units=case_b['ipus'], method='knn', k=4
)
np.testing.assert_almost_equal(
calculate_metric("convergence", time_series_a, time_series_b),
0.00022443279925280575,
)
def test_calculate_synchrony_oposites(self):
case_a = self.cases['long_100-200-300_x2']
case_b = self.cases['long_300-200-100_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
time_series_b = TimeSeries(
feature='F0_MAX', interpausal_units=case_b['ipus'], method='knn', k=4
)
np.testing.assert_almost_equal(
calculate_metric(
"synchrony",
time_series_a,
time_series_b,
),
-0.925558204113492,
)
def test_calculate_synchrony_w_deltas_bigger_than_interval(self):
case_a = self.cases['long_100-200-300_x2']
case_b = self.cases['long_300-200-100_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
time_series_b = TimeSeries(
feature='F0_MAX', interpausal_units=case_b['ipus'], method='knn', k=4
)
self.assertRaises(
ValueError,
calculate_metric,
"synchrony",
time_series_a,
time_series_b,
synchrony_deltas=[100.0],
)
def test_calculate_synchrony_w_different_methods_is_similar(self):
case_a = self.cases['long_100-200-300_x2']
case_b = self.cases['long_300-200-100_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
time_series_b = TimeSeries(
feature='F0_MAX', interpausal_units=case_b['ipus'], method='knn', k=4
)
np.testing.assert_almost_equal(
calculate_metric(
"synchrony",
time_series_a,
time_series_b,
integration_method="montecarlo",
),
calculate_metric(
"synchrony",
time_series_a,
time_series_b,
integration_method="trapz",
),
decimal=4,
)
def test_predict_time_series_with_invalid_input_raises_exception(self):
case_a = self.cases['long_100-200-300_x2']
time_series_a = TimeSeries(
feature='F0_MAX', interpausal_units=case_a['ipus'], method='knn', k=4
)
invalid_inputs = [1, np.array([[1]])]
for invalid_input in invalid_inputs:
self.assertRaises(
ValueError,
time_series_a.predict,
invalid_input,
)
def test_calculate_knn_time_series_warns_none_feature_value(self):
case = self.cases['ipu_with_None_value']
self.assertWarns(
Warning,
TimeSeries,
feature='F0_MAX',
interpausal_units=case['ipus'],
method='knn',
k=3,
)
def test_calculate_knn_time_series_warns_nan_feature_value(self):
case = self.cases['ipu_with_nan_value']
self.assertWarns(
Warning,
TimeSeries,
feature='F0_MAX',
interpausal_units=case['ipus'],
method='knn',
k=3,
)
def test_predict_interval_over_time_series_from_unordered_ipus(self):
case = self.cases['unordered']
self.assertWarns(
Warning,
TimeSeries,
feature='F0_MAX',
interpausal_units=case['ipus'],
method='knn',
k=3,
)