-
Notifications
You must be signed in to change notification settings - Fork 1
/
3-run_art.py
180 lines (136 loc) · 7.39 KB
/
3-run_art.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
import os
from art_utils import med_vs_lowres, high_vs_lowres, per_vs_cnn
file_dir = os.path.abspath(os.path.dirname(__file__))
if __name__ == "__main__":
# number of data shuffles for approximate randomization testing
N = 100000
log_dir = os.path.join(file_dir, 'results', 'art')
if not os.path.exists(log_dir):
os.makedirs(log_dir)
data_dir = os.path.join(file_dir, 'Featurize', 'featurized')
models_dir = os.path.join(file_dir, 'results', 'models')
highres_words = os.path.join(data_dir, 'words', 'hires_test')
medres_words = os.path.join(data_dir, 'words', 'medres_test')
lowres_words = os.path.join(data_dir, 'words', 'lores_test')
highres_gender = os.path.join(data_dir, 'gender', 'hires_test')
medres_gender = os.path.join(data_dir, 'gender', 'medres_test')
lowres_gender = os.path.join(data_dir, 'gender', 'lores_test')
params = dict()
params['highres_id'] = 0
params['medres_id'] = 1
params['lowres_id'] = 2
params['bonferroni'] = 18
# run ART comparisons for word recognition
###########################################################################
params['lowres_path'] = lowres_words
params['medres_path'] = medres_words
params['highres_path'] = highres_words
params['logpath'] = os.path.join(log_dir, 'words_cd_cnn.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_cnn')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_cd_per.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_per')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_cd_cnn_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_cnn_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_cd_cnn_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_cnn_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_cd_per_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_per_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_cd_per_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_cd_per_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_cnn.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_cnn')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_per.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_per')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_cnn_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_cnn_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_cnn_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_cnn_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_per_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_per_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'words_pd_per_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'words_pd_per_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'high_vs_lowres_words_per.csv')
params['highres_model_dir'] = os.path.join(models_dir, 'words_pd_per')
params['lowres_model_dir'] = os.path.join(models_dir, 'words_cd_per')
high_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'high_vs_lowres_words_cnn.csv')
params['highres_model_dir'] = os.path.join(models_dir, 'words_pd_cnn')
params['lowres_model_dir'] = os.path.join(models_dir, 'words_cd_cnn')
high_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'per_vs_cnn_words_pd.csv')
params['per_dir'] = os.path.join(models_dir, 'words_pd_per')
params['cnn_dir'] = os.path.join(models_dir, 'words_pd_cnn')
per_vs_cnn(params, N)
params['logpath'] = os.path.join(log_dir, 'per_vs_cnn_words_cd.csv')
params['per_dir'] = os.path.join(models_dir, 'words_cd_per')
params['cnn_dir'] = os.path.join(models_dir, 'words_cd_cnn')
per_vs_cnn(params, N)
# run ART comparisons for gender recognition
###########################################################################
params['lowres_path'] = lowres_gender
params['medres_path'] = medres_gender
params['highres_path'] = highres_gender
params['logpath'] = os.path.join(log_dir, 'gender_cd_cnn.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_cnn')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_cd_per.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_per')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_cd_cnn_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_cnn_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_cd_cnn_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_cnn_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_cd_per_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_per_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_cd_per_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_cd_per_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_cnn.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_cnn')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_per.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_per')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_cnn_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_cnn_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_cnn_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_cnn_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_per_0ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_per_0ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'gender_pd_per_1ep.csv')
params['model_dir'] = os.path.join(models_dir, 'gender_pd_per_1ep')
med_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'high_vs_lowres_gender_per.csv')
params['highres_model_dir'] = os.path.join(models_dir, 'gender_pd_per')
params['lowres_model_dir'] = os.path.join(models_dir, 'gender_cd_per')
high_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'high_vs_lowres_gender_cnn.csv')
params['highres_model_dir'] = os.path.join(models_dir, 'gender_pd_cnn')
params['lowres_model_dir'] = os.path.join(models_dir, 'gender_cd_cnn')
high_vs_lowres(params, N)
params['logpath'] = os.path.join(log_dir, 'per_vs_cnn_gender_pd.csv')
params['per_dir'] = os.path.join(models_dir, 'gender_pd_per')
params['cnn_dir'] = os.path.join(models_dir, 'gender_pd_cnn')
per_vs_cnn(params, N)
params['logpath'] = os.path.join(log_dir, 'per_vs_cnn_gender_cd.csv')
params['per_dir'] = os.path.join(models_dir, 'gender_cd_per')
params['cnn_dir'] = os.path.join(models_dir, 'gender_cd_cnn')
per_vs_cnn(params, N)