-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathSTM13_plot_averagedSTM.py
397 lines (327 loc) · 13.6 KB
/
STM13_plot_averagedSTM.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
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu Sep 5 22:54:42 2024
@author: andrewchang
"""
import numpy as np
import pandas as pd
import keras
from keras import layers
import keras_tuner as kt
import datetime
import os
import tensorflow as tf
import gc
import glob
import scipy.io
os.environ["KERAS_BACKEND"] = "tensorflow"
# %% load data
# def prepData_plotSTM():
# % load STM data
corpus_speech_list = ['BibleTTS/akuapem-twi',
'BibleTTS/asante-twi',
'BibleTTS/ewe',
'BibleTTS/hausa',
'BibleTTS/lingala',
'BibleTTS/yoruba',
'Buckeye',
'EUROM',
'HiltonMoser2022_speech',
'LibriSpeech',
# 'LibriVox',
'MediaSpeech/AR',
'MediaSpeech/ES',
'MediaSpeech/FR',
'MediaSpeech/TR',
'MozillaCommonVoice/ab',
'MozillaCommonVoice/ar',
'MozillaCommonVoice/ba',
'MozillaCommonVoice/be',
'MozillaCommonVoice/bg',
'MozillaCommonVoice/bn',
'MozillaCommonVoice/br',
'MozillaCommonVoice/ca',
'MozillaCommonVoice/ckb',
'MozillaCommonVoice/cnh',
'MozillaCommonVoice/cs',
'MozillaCommonVoice/cv',
'MozillaCommonVoice/cy',
'MozillaCommonVoice/da',
'MozillaCommonVoice/de',
'MozillaCommonVoice/dv',
'MozillaCommonVoice/el',
'MozillaCommonVoice/en',
'MozillaCommonVoice/eo',
'MozillaCommonVoice/es',
'MozillaCommonVoice/et',
'MozillaCommonVoice/eu',
'MozillaCommonVoice/fa',
'MozillaCommonVoice/fi',
'MozillaCommonVoice/fr',
'MozillaCommonVoice/fy-NL',
'MozillaCommonVoice/ga-IE',
'MozillaCommonVoice/gl',
'MozillaCommonVoice/gn',
'MozillaCommonVoice/hi',
'MozillaCommonVoice/hu',
'MozillaCommonVoice/hy-AM',
'MozillaCommonVoice/id',
'MozillaCommonVoice/ig',
'MozillaCommonVoice/it',
'MozillaCommonVoice/ja',
'MozillaCommonVoice/ka',
'MozillaCommonVoice/kab',
'MozillaCommonVoice/kk',
'MozillaCommonVoice/kmr',
'MozillaCommonVoice/ky',
'MozillaCommonVoice/lg',
'MozillaCommonVoice/lt',
'MozillaCommonVoice/ltg',
'MozillaCommonVoice/lv',
'MozillaCommonVoice/mhr',
'MozillaCommonVoice/ml',
'MozillaCommonVoice/mn',
'MozillaCommonVoice/mt',
'MozillaCommonVoice/nan-tw',
'MozillaCommonVoice/nl',
'MozillaCommonVoice/oc',
'MozillaCommonVoice/or',
'MozillaCommonVoice/pl',
'MozillaCommonVoice/pt',
'MozillaCommonVoice/ro',
'MozillaCommonVoice/ru',
'MozillaCommonVoice/rw',
'MozillaCommonVoice/sr',
'MozillaCommonVoice/sv-SE',
'MozillaCommonVoice/sw',
'MozillaCommonVoice/ta',
'MozillaCommonVoice/th',
'MozillaCommonVoice/tr',
'MozillaCommonVoice/tt',
'MozillaCommonVoice/ug',
'MozillaCommonVoice/uk',
'MozillaCommonVoice/ur',
'MozillaCommonVoice/uz',
'MozillaCommonVoice/vi',
'MozillaCommonVoice/yo',
'MozillaCommonVoice/yue',
'MozillaCommonVoice/zh-CN',
'MozillaCommonVoice/zh-TW',
'primewords_chinese',
'room_reader',
'SpeechClarity',
'TAT-Vol2',
'thchs30',
'TIMIT',
'TTS_Javanese',
'zeroth_korean'
]
corpus_music_list = [
'IRMAS',
'Albouy2020Science',
# 'CD', # exclude CDs due to open source concern
'GarlandEncyclopedia',
'fma_large',
'ismir04_genre',
'MTG-Jamendo',
'HiltonMoser2022_song',
'NHS2',
'MagnaTagATune'
]
corpus_env_list = ['SONYC', 'MacaulayLibrary']
# sort the corpora lists to make sure the order is replicable
corpus_speech_list.sort()
corpus_music_list.sort()
corpus_env_list.sort()
corpus_list_all = corpus_speech_list+corpus_music_list+corpus_env_list
for corp in corpus_list_all:
filename = 'STM_output/corpSTMnpy/'+corp.replace('/', '-')+'_STMall.npy'
if 'STM_all' not in locals():
STM_all = np.load(filename)
else:
STM_all = np.vstack((STM_all, np.load(filename)))
print(filename)
# % load meta data
speech_corp_df1 = pd.read_csv('train_test_split/speech1_10folds_speakerGroupFold.csv',index_col=0)
speech_corp_df2 = pd.read_csv('train_test_split/speech2_10folds_speakerGroupFold.csv',index_col=0)
music_corp_df = pd.read_csv('train_test_split/music_10folds_speakerGroupFold.csv',index_col=0)
df_SONYC = pd.read_csv('train_test_split/env_10folds_speakerGroupFold.csv',index_col=0)
all_corp_df = pd.concat([speech_corp_df1, speech_corp_df2, music_corp_df, df_SONYC], ignore_index=True)
target = all_corp_df['corpus_type']
data_split = all_corp_df['10fold_labels']
target.replace({
'speech: non-tonal':0,
'speech: tonal':1,
'music: vocal':2,
'music: non-vocal':3,
'env: urban':4,
'env: wildlife':5,
},
inplace=True)
mean_STM_list = []
for t in range(len(target.unique())):
mean_STM_list.append(np.mean(STM_all[target==t,:], axis = 0).reshape(20,121))
# %% Cohen's d on STM
def cohenD_STM(d1, d2):
mean1 = np.mean(d1, axis=0)
mean2 = np.mean(d2, axis=0)
len1 = d1.shape[0]
len2 = d2.shape[0]
s1 = np.std(d1, axis=0, ddof=1)
s2 = np.std(d1, axis=0, ddof=1)
sp = np.sqrt(((len1-1)*s1**2 + (len2-1)*s2**2)/(len1+len2-2))
cohenD = (mean1-mean2)/sp
return cohenD
cohenD_withinSpeech = cohenD_STM(STM_all[target==0,:], STM_all[target==1,:]).reshape(20,121)
cohenD_withinMusic = cohenD_STM(STM_all[target==2,:], STM_all[target==3,:]).reshape(20,121)
cohenD_withinEnv = cohenD_STM(STM_all[target==4,:], STM_all[target==5,:]).reshape(20,121)
cohenD_SpeechMusic = cohenD_STM(STM_all[target.isin([0,1]),:], STM_all[target.isin([2,3]),:]).reshape(20,121)
cohenD_MusicEnv = cohenD_STM(STM_all[target.isin([2,3]),:], STM_all[target.isin([4,5]),:]).reshape(20,121)
cohenD_SpeechEnv = cohenD_STM(STM_all[target.isin([0,1]),:], STM_all[target.isin([4,5]),:]).reshape(20,121)
# %% do t-test (ver slow!)
# from scipy.stats import ttest_ind
# result_withinSpeech = ttest_ind(STM_all[target==0,:], STM_all[target==1,:], axis=0, equal_var=False)
# result_withinMusic = ttest_ind(STM_all[target==2,:], STM_all[target==3,:], axis=0, equal_var=False)
# result_withinEnv = ttest_ind(STM_all[target==4,:], STM_all[target==5,:], axis=0, equal_var=False)
# result_withinSpeech = ttest_ind(STM_all[target==0,:], STM_all[target==1,:], axis=0, equal_var=False)
# %% get the STM axes
df = pd.read_csv('metaTables/metaData_BibleTTS-hausa.csv',index_col=0)
survey_file = df['mat_filename'][0].replace('MATs','Survey').replace('mat_wl4','params').replace('MS2024','Params')
survey_data = scipy.io.loadmat(survey_file)
x_axis = survey_data['Params'][0]['x_axis'][0][0]
y_axis = survey_data['Params'][0]['y_axis'][0][0]
xrange = [-15, 15]
yrange = [0,7.2]
x_ds_factor = 1
y_ds_factor = 2
xmin = np.argmin(np.abs(x_axis - xrange[0]))
xmax = np.argmin(np.abs(x_axis - xrange[1]))
ymin = np.argmin(np.abs(y_axis - yrange[0]))
ymax = np.argmin(np.abs(y_axis - yrange[1]))
x_axis_small = x_axis[xmin:xmax+1:x_ds_factor]
y_axis_small = y_axis[ymin:ymax+1:y_ds_factor]
# %% find max location (buggy!!)
a = np.argmax(STM_all[target==0,:], axis=1)
a = STM_all[0,:].reshape(20,121)
def normalize_STM(STM_all, x_axis_small, y_axis_small):
x_normalizer = np.tile(np.sqrt(abs(np.tile(x_axis_small, (len(y_axis_small),1)))).flatten(),(len(STM_all), 1))
y_normalizer = np.tile(np.sqrt(abs(np.tile(y_axis_small, (len(x_axis_small),1)).T)).flatten(),(len(STM_all), 1))
return STM_all/x_normalizer/y_normalizer
STM_all_norm = normalize_STM(STM_all, x_axis_small, y_axis_small)
from numpy import inf
STM_all_norm[STM_all_norm==inf]=0
STM_max_loc = np.argmax(STM_all_norm, axis=1)
def mark_peak(STM_max_loc, target, cond):
zeros = np.zeros(2420)
STM_max_loc = STM_max_loc[target==cond]
for n in STM_max_loc:
zeros[n]+=1
return zeros
max_1 = mark_peak(STM_max_loc, target, 3)
# %% plot mean STMs
import matplotlib.pyplot as plt
import matplotlib
import matplotlib.colors as mcolors
with plt.style.context('seaborn-v0_8-poster'):
plt.rcParams['figure.dpi'] = 300
plt.rcParams['savefig.dpi'] = 300
# Create a figure and a grid of subplots
fig, ax = plt.subplots(2, 3, figsize=(15, 7))
extent = [x_axis_small.min(), x_axis_small.max(), y_axis_small.min(), y_axis_small.max()]
norm = mcolors.LogNorm(vmin=0.7*1e-1, vmax=1)
cmap='jet'
# Plot the first image in the top-left subplot
im0 = ax[0, 0].imshow(mean_STM_list[0], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 0].set_title('Speech: nontonal')
# ax[0, 0].set_xlabel('temporal modulation (Hz)')
# ax[0, 0].set_ylabel('spectral modulation (cyc/oct)')
im1 = ax[1, 0].imshow(mean_STM_list[1], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 0].set_title('Speech: tonal')
ax[1, 0].set_xlabel('temporal modulation (Hz)')
ax[1, 0].set_ylabel('spectral modulation (cyc/oct)')
im2 = ax[0, 1].imshow(mean_STM_list[2], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 1].set_title('Music: vocal')
# ax[0, 1].set_xlabel('temporal modulation (Hz)')
# ax[0, 1].set_ylabel('spectral modulation (cyc/oct)')
im3 = ax[1, 1].imshow(mean_STM_list[3], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 1].set_title('Music: nonvocal')
# ax[1, 1].set_xlabel('temporal modulation (Hz)')
# ax[1, 1].set_ylabel('spectral modulation (cyc/oct)')
im4 = ax[0, 2].imshow(mean_STM_list[4], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 2].set_title('Env: urban')
# ax[0, 2].set_xlabel('temporal modulation (Hz)')
# ax[0, 2].set_ylabel('spectral modulation (cyc/oct)')
im5 = ax[1, 2].imshow(mean_STM_list[5], aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 2].set_title('Env: wildlife')
# ax[1, 2].set_xlabel('temporal modulation (Hz)')
# ax[1, 2].set_ylabel('spectral modulation (cyc/oct)')
cbar_ax = fig.add_axes([0.9, 0.2, 0.01, 0.6])
cbar = plt.colorbar(im1, cax=cbar_ax)
# Add a single colorbar on the top side
# cbar = fig.colorbar(im5, ax=cbar_ax, orientation='horizontal', fraction=0.04, pad=0.2)
cbar.set_label('normalized power')
# cbar.set_ticks(ticks=[0.1, 0.5, 1], labels=['0.1', '0.5', '1'])
cbar.set_ticks([0.1, 0.5, 1])
# cbar.set_ticklabels(['0.1', '0.5', '1'])
# Adjust layout to make room for the colorbar
plt.tight_layout(rect=[0, 0, 0.9, 1]) # Adjust the rect parameter to fit the colorbar
# plt.tight_layout()
# plt.tight_layout()
plt.show()
fig.savefig('categorical_mean_STM20240906.png')
# %% plot STMs Cohen's d
import matplotlib.pyplot as plt
import matplotlib
import matplotlib.colors as mcolors
with plt.style.context('seaborn-v0_8-poster'):
plt.rcParams['figure.dpi'] = 300
plt.rcParams['savefig.dpi'] = 300
# Create a figure and a grid of subplots
fig, ax = plt.subplots(2, 3, figsize=(15, 7))
extent = [x_axis_small.min(), x_axis_small.max(), y_axis_small.min(), y_axis_small.max()]
# norm = mcolors.NoNorm(vmin=np.nanmin([cohenD_withinSpeech, cohenD_withinMusic, cohenD_withinEnv, cohenD_SpeechMusic, cohenD_MusicEnv, cohenD_SpeechEnv]),
# vmax=np.nanmax([cohenD_withinSpeech, cohenD_withinMusic, cohenD_withinEnv, cohenD_SpeechMusic, cohenD_MusicEnv, cohenD_SpeechEnv]))
# norm = mcolors.NoNorm(vmin=-5, vmax=5)
# norm = mcolors.SymLogNorm(linthresh=0.5, vmin=-10.0, vmax=10.0, base=10, clip=True)
norm = mcolors.AsinhNorm(linear_width=2, vmin=-5.0, vmax=5.0, clip=True)
cmap='bwr'
# Plot the first image in the top-left subplot
im0 = ax[0, 0].imshow(cohenD_withinSpeech, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 0].set_title('Speech: nontonal - tonal')
# ax[0, 0].set_xlabel('temporal modulation (Hz)')
# ax[0, 0].set_ylabel('spectral modulation (cyc/oct)')
im1 = ax[1, 0].imshow(cohenD_SpeechMusic, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 0].set_title('Speech - Music')
ax[1, 0].set_xlabel('temporal modulation (Hz)')
ax[1, 0].set_ylabel('spectral modulation (cyc/oct)')
im2 = ax[0, 1].imshow(cohenD_withinMusic, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 1].set_title('Music: vocal - nonvocal')
# ax[0, 1].set_xlabel('temporal modulation (Hz)')
# ax[0, 1].set_ylabel('spectral modulation (cyc/oct)')
im3 = ax[1, 1].imshow(cohenD_SpeechEnv, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 1].set_title('Speech - Env')
# ax[1, 1].set_xlabel('temporal modulation (Hz)')
# ax[1, 1].set_ylabel('spectral modulation (cyc/oct)')
im4 = ax[0, 2].imshow(cohenD_withinEnv, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[0, 2].set_title('Env: urban - wildlife')
# ax[0, 2].set_xlabel('temporal modulation (Hz)')
# ax[0, 2].set_ylabel('spectral modulation (cyc/oct)')
im5 = ax[1, 2].imshow(cohenD_MusicEnv, aspect=2.5, origin='lower', extent=extent, norm=norm, cmap=cmap)
ax[1, 2].set_title('Music - Env')
# ax[1, 2].set_xlabel('temporal modulation (Hz)')
# ax[1, 2].set_ylabel('spectral modulation (cyc/oct)')
cbar_ax = fig.add_axes([0.9, 0.2, 0.01, 0.6])
cbar = plt.colorbar(im1, cax=cbar_ax)
# Add a single colorbar on the top side
# cbar = fig.colorbar(im5, ax=cbar_ax, orientation='horizontal', fraction=0.04, pad=0.2)
cbar.set_label("Cohen's d")
cbar.set_ticks(ticks=[-5, -1, 0, 1, 5], labels=['-5', '-1', '0', '1', '5'])
# cbar.set_ticks([-2.5, -1, 0, 1, 2.5])
# cbar.set_ticklabels(['0.1', '0.5', '1'])
# Adjust layout to make room for the colorbar
plt.tight_layout(rect=[0, 0, 0.9, 1]) # Adjust the rect parameter to fit the colorbar
# plt.tight_layout()
# plt.tight_layout()
plt.show()
fig.savefig('categorical_CohenD_STM20240911.png')