-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathexp_generalization.py
226 lines (185 loc) · 9.15 KB
/
exp_generalization.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
from genericpath import exists
import numpy as np
import models.train_and_evaluate
import json
import copy
import os
import pandas as pd
import figure_cv_bacc
import figure_cm
import figure_auc_roc_curve
import models.mn
import models.null
import concurrent.futures
def load_spec(path):
with open(path, 'r') as f:
return json.load(f)
def main():
mn_spec = load_spec("cfgs/smf_mn_model.json")
mn_spec['target_col'] = 'is_viable'
mn_spec['batch_size_p'] = 0.1
model_output_paths = train_smf_model(mn_spec, "../results/exp_generalization/smf_model")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_pombe_smf.feather", "../results/exp_generalization/s-mn_pombe.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_human_smf.feather", "../results/exp_generalization/s-mn_human.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_dro_smf.feather", "../results/exp_generalization/s-mn_dro.json")
null_spec = { 'target_col' : 'is_viable', 'class' : 'null' }
model_output_paths = train_smf_model(null_spec, "../results/exp_generalization/s-null_model")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_pombe_smf.feather", "../results/exp_generalization/s-null_pombe.json")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_human_smf.feather", "../results/exp_generalization/s-null_human.json")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_dro_smf.feather", "../results/exp_generalization/s-null_dro.json")
generate_smf_figures('pombe')
generate_smf_figures('human')
generate_smf_figures('dro')
mn_spec = load_spec("cfgs/gi_mn_model.json")
mn_spec['target_col'] = 'is_neutral'
model_output_paths = train_gi_model(mn_spec, "../results/exp_generalization/gi_model")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_pombe_gi_mn.feather", "../results/exp_generalization/d-mn_pombe.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_human_gi_mn.feather", "../results/exp_generalization/d-mn_human.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_dro_gi_mn.feather", "../results/exp_generalization/d-mn_dro.json")
mn_spec = load_spec("cfgs/gi_mn_model.json")
mn_spec['target_col'] = 'is_neutral'
mn_spec['features'].remove('sgo-')
model_output_paths = train_gi_model(mn_spec, "../results/exp_generalization/gi_model_no_sgo")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_pombe_gi_mn.feather", "../results/exp_generalization/d-mn_no_sgo_pombe.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_human_gi_mn.feather", "../results/exp_generalization/d-mn_no_sgo_human.json")
evaluate(mn_spec, model_output_paths, "../generated-data/dataset_dro_gi_mn.feather", "../results/exp_generalization/d-mn_no_sgo_dro.json")
null_spec = { 'target_col' : 'is_neutral', 'class' : 'null' }
model_output_paths = train_gi_model(null_spec, "../results/exp_generalization/d-null")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_pombe_gi_mn.feather", "../results/exp_generalization/d-null_pombe.json")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_human_gi_mn.feather", "../results/exp_generalization/d-null_human.json")
evaluate(null_spec, model_output_paths, "../generated-data/dataset_dro_gi_mn.feather", "../results/exp_generalization/d-null_dro.json")
generate_gi_figures('pombe')
generate_gi_figures('human')
generate_gi_figures('dro')
def train_smf_model(mn_spec, trained_model_path):
df = pd.read_feather('../generated-data/dataset_yeast_smf.feather')
d = np.load("../generated-data/splits/dataset_yeast_smf.npz", allow_pickle=True)
splits = d['splits']
n_reps = d['reps']
n_folds = d['folds']
split_ids = [i * n_folds for i in range(n_reps)]
return train(mn_spec, df, splits, split_ids, trained_model_path)
def train_gi_model(mn_spec, trained_model_path):
df = pd.read_feather('../generated-data/dataset_yeast_gi_hybrid_mn.feather')
d = np.load("../generated-data/splits/dataset_yeast_gi_hybrid.npz",
allow_pickle=True)
splits = d['splits']
n_reps = d['reps']
n_folds = d['folds']
split_ids = [i * n_folds for i in range(n_reps)]
return train(mn_spec, df, splits, split_ids, trained_model_path)
def train(model_spec, df, splits, split_ids, output_path):
os.makedirs(output_path, exist_ok=True)
model_spec['verbose'] = False
futures = []
model_output_paths = []
with concurrent.futures.ThreadPoolExecutor(max_workers=len(split_ids)) as executor:
for i, split_id in enumerate(split_ids):
train_df, valid_df, _ = models.common.get_dfs(df, splits[split_id], train_ids=[0, 1, 3], valid_ids=[2], test_ids=[0])
model_output_path = os.path.join(output_path, 'model_%d.npz' % i)
model_output_paths.append(model_output_path)
futures.append(executor.submit(_train_model, model_spec, train_df, valid_df, model_output_path))
concurrent.futures.wait(futures)
return model_output_paths
def _train_model(model_spec, train_df, valid_df, output_path):
model_spec = copy.deepcopy(model_spec)
if model_spec['class'] == 'mn':
model = models.mn.MnModel(model_spec)
else:
model = models.null.NullModel(model_spec)
model.train(train_df, valid_df)
model.save(output_path)
def evaluate(model_spec, model_output_paths, test_path, results_path):
model_class = models.mn.MnModel if model_spec['class'] == 'mn' else models.null.NullModel
test_df = pd.read_feather(test_path)
preds = []
for model_output_path in model_output_paths:
model = model_class.load(model_output_path)
preds.append(model.predict(test_df, training_norm=False))
preds = np.array(preds)
preds = np.mean(preds, axis=0)
target_col = model_spec['target_col']
r = models.common.evaluate(np.array(test_df[target_col]), preds)
with open(results_path, "w") as f:
json.dump({ "model_spec" : model_spec, "results" : [r] }, f, indent=4)
def generate_smf_figures(org):
output_dir = "../results/exp_generalization/figures/smf/%s" % org
os.makedirs(output_dir, exist_ok=True)
spec = {
"models": [
{
"title": "S-MN",
"color": "#3A90FF",
"name" : "s-mn_%s" % org,
"results_path" : "../results/exp_generalization/s-mn_%s.json" % org
},
{
"title": "Null",
"color": "#c9c9c9",
"star_color": "grey",
"cm_color": "grey",
"name" : "s-null_%s" % org,
"results_path" : "../results/exp_generalization/s-null_%s.json" % org
}
],
"classes": [
"Lethal",
"Viable"
],
"short_classes": [
"L",
"V"
],
"ylim" : [0,1],
"aspect" : 1
}
figure_cv_bacc.generate_figures(spec, "../results/exp_generalization", os.path.join(output_dir, 'overall_bacc.png'))
for model in spec['models']:
figure_cm.plot_cm(model['results_path'], model['color'], spec['short_classes'], os.path.join(output_dir, "cm_%s.png" % model['name']))
for i in range(len(spec['classes'])):
figure_auc_roc_curve.plot_auc_roc_curves(spec, i, os.path.join(output_dir, "auc_roc%s.png" % spec["short_classes"][i]))
def generate_gi_figures(org):
output_dir = "../results/exp_generalization/figures/gi/%s" % org
os.makedirs(output_dir, exist_ok=True)
spec = {
"models": [
{
"title": "D-MN",
"color": "#3A90FF",
"name" : "d-mn_%s" % org,
"results_path" : "../results/exp_generalization/d-mn_%s.json" % org
},
{
"title": "D-MN No sGO",
"color": "#38fffc",
"name" : "d-mn_no_sgo_%s" % org,
"fsize" : 50,
"results_path" : "../results/exp_generalization/d-mn_no_sgo_%s.json" % org
},
{
"title": "Null",
"color": "#c9c9c9",
"star_color": "grey",
"cm_color": "grey",
"name" : "d-null_%s" % org,
"results_path" : "../results/exp_generalization/d-null_%s.json" % org
}
],
"classes": [
"Interacting",
"Neutral"
],
"short_classes": [
"I",
"N"
],
"ylim" : [0,1],
"aspect" : 1
}
figure_cv_bacc.generate_figures(spec, "../results/exp_generalization", os.path.join(output_dir, 'overall_bacc.png'))
for model in spec['models']:
figure_cm.plot_cm(model['results_path'], model['color'], spec['short_classes'], os.path.join(output_dir, "cm_%s.png" % model['name']))
for i in range(len(spec['classes'])):
figure_auc_roc_curve.plot_auc_roc_curves(spec, i, os.path.join(output_dir, "auc_roc%s.png" % spec["short_classes"][i]))
if __name__ == "__main__":
main()