-
Notifications
You must be signed in to change notification settings - Fork 1
/
baseline_run.py
242 lines (228 loc) · 9.41 KB
/
baseline_run.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
from VSBO_class import *
import argparse
parser = argparse.ArgumentParser('VS-BO')
parser.add_argument('--obj_func', type=str)
parser.add_argument('--method', type=str)
args = parser.parse_args()
def makedirs(dirname):
if not os.path.exists(dirname):
os.makedirs(dirname)
if args.obj_func=="Branin":
### Branin test with D=50 and d_{e}=[2,2,2]
object_dim = 50
object_bounds = torch.cat([generate_branin_bounds(2),generate_branin_bounds(2),generate_branin_bounds(object_dim-4)],dim=1)
object_func = Combine_func
function_kwargs = {'func_set':[Branin_hd,Branin_hd,Branin_hd],'var_num_set':[2,2,2],'coeff_set':[1,0.1,0.01]}
total_budget = 210
total_time = 600
total_cput = 9000
elif args.obj_func=="Hartmann6":
### Hartmann6 test with D=50 and d_{e}=[6,6,6]
object_dim = 50
object_bounds = generate_hartmann_bounds(object_dim)
object_func = Combine_func
function_kwargs = {'func_set':[Hartmann6,Hartmann6,Hartmann6],'var_num_set':[6,6,6],'coeff_set':[1,0.1,0.01]}
total_budget = 210
total_time = 400
total_cput = 7000
elif args.obj_func=="StyblinskiTang4":
### StyblinskiTang4 test with D=50 and d_{e}=[4,4,4]
object_dim = 50
object_bounds = generate_StyblinskiTang_bounds(object_dim)
object_func = Combine_func
function_kwargs = {'func_set':[StyblinskiTang4_hd,StyblinskiTang4_hd,StyblinskiTang4_hd],'var_num_set':[4,4,4],'coeff_set':[1,0.1,0.01]}
total_budget = 210
total_time = 2000
total_cput = 30000
elif args.obj_func=="rover":
from rover_test_utils import *
object_dim = 60
object_bounds = generate_rover_bounds()
object_func = rover_func
total_budget = 210
total_time = 2000
total_cput = -1
function_kwargs={}
elif args.obj_func=="mopta":
from mopta_utils import *
object_dim = 124
object_bounds = generate_mopta_bounds()
object_func = mopta_func
total_budget = 210
total_time = 3000
total_cput = -1
function_kwargs={}
output_path = "./"+args.obj_func+"/"+args.method+"/"
makedirs(output_path)
init_samples = 5
if args.method == "vanillaBO":
for test_id in range(1,21):
BO_instance = BOtorch(object_dim,object_func,obj_func_kwargs=function_kwargs,bounds=object_bounds)
BO_instance.data_initialize()
Times = []
T_process = []
t0 = time.time()
t1 = time.process_time()
iter_num = 0
while (time.time() -t0 < total_time or iter_num < total_budget or time.process_time()-t1 < total_cput):
iter_num+=1
try:
BO_instance.GP_fitting(GP_Matern)
BO_instance.BO_acq_optim(optim_method='LBFGS')
BO_instance.data_update()
Times.append(time.time()-t0)
T_process.append(time.process_time()-t1)
except ValueError as e:
if(e.args[0]=='Too many cov mat singular!'):
BO_instance.erase_last_instance()
iter_num-=1
continue
else:
raise ValueError(e.args[0])
if(iter_num%10==0):
print(
f"Epoch {iter_num:>3} "
f"Best value: {torch.max(BO_instance.Y).item():>4.3f}"
)
np.save(output_path+"X_"+str(test_id)+".npy",BO_instance.X.numpy())
np.save(output_path+"Y_"+str(test_id)+".npy",BO_instance.Y.numpy())
np.save(output_path+"Time_"+str(test_id)+".npy",np.array(Times))
np.save(output_path+"Time_process_"+str(test_id)+".npy",np.array(T_process))
elif args.method == "rembo":
from rembo_utils import *
embed_dim = 6
for test_id in range(1,21):
opt = REMBOOptimizer(GP_model=GP_Matern,initial_random_samples=init_samples,n_dims=object_dim,n_embedding_dims=embed_dim)
opt.model_initialise(object_bounds.T.numpy())
opt.y_ = object_func(opt.X_,**function_kwargs)
_,_ = opt.fit_model()
Times = []
T_process = []
t0 = time.time()
t1 = time.process_time()
iter_num = 0
#while time.time() - t0 < total_time:
while (time.time() -t0 < total_time or iter_num < total_budget or time.process_time()-t1 < total_cput):
iter_num+=1
x_new = opt.select_query_point()
#x_new[:,integer_f_list] = torch.round(x_new[:,integer_f_list])
y_new = object_func(x_new,**function_kwargs)
_,_ = opt.update(x_new,y_new)
Times.append(time.time()-t0)
T_process.append(time.process_time()-t1)
if(iter_num%10==0):
print(
f"Epoch {iter_num:>3} "
f"Best value: {opt.best_value().item():>4.3f}"
)
np.save(output_path + "X_"+str(test_id)+".npy",opt.X_.numpy())
np.save(output_path + "Y_"+str(test_id)+".npy",opt.y_.numpy())
np.save(output_path + "Time_"+str(test_id)+".npy",np.array(Times))
np.save(output_path + "Time_process_"+str(test_id)+".npy",np.array(T_process))
elif args.method=="rembo_interleave":
from rembo_utils import *
embed_dim = 6
interleaved_cycle = 4
for test_id in range(1,21):
opt = InterleavedREMBOOptimizer(interleaved_runs=interleaved_cycle,random_state=test_id,GP_model=GP_Matern,n_dims=object_dim,n_embedding_dims=embed_dim)
for i in range(interleaved_cycle):
opt.rembos[i].model_initialise(object_bounds.T.numpy())
opt.rembos[i].y_ = object_func(opt.rembos[i].X_,**function_kwargs)
_,_ = opt.rembos[i].fit_model()
#opt.model_initialise(object_bounds.T.numpy())
#opt.y_ = object_func(opt.X_)
opt.get_cycle()
#_,_ = opt.fit_model()
Times = []
T_process = []
t0 = time.time()
t1 = time.process_time()
iter_num = 0
while (time.time() -t0 < total_time or iter_num < total_budget or time.process_time()-t1 < total_cput):
iter_num+=1
x_new = opt.select_query_point()
#x_new[:,integer_f_list] = torch.round(x_new[:,integer_f_list])
y_new = object_func(x_new,**function_kwargs)
opt.update(x_new,y_new)
Times.append(time.time()-t0)
T_process.append(time.process_time()-t1)
if(iter_num%10==0):
print(
f"Epoch {iter_num:>3} "
f"Best value: {opt.best_value().item():>4.3f}"
)
np.save(output_path + "X_"+str(test_id)+".npy",opt.X_.numpy())
np.save(output_path + "Y_"+str(test_id)+".npy",opt.y_.numpy())
np.save(output_path + "Time_"+str(test_id)+".npy",np.array(Times))
np.save(output_path + "Time_process_"+str(test_id)+".npy",np.array(T_process))
### Alebo Hesbo pipeline
from alebo_hesbo_utils import *
#Times = []
#T_process = []
#wct = 0
#cput = 0
#y_max = -100
#embed_dim = 6
parameters = convert_botorch_bounds_to_ax(object_bounds)
def object_func_ax_wrapper(parameterization):
global Times
global T_process
global y_max
#global function_kwargs
#pdb.set_trace()
Times.append(time.time()-wct)
T_process.append(time.process_time()-cput)
dim = len(parameterization)
x = torch.tensor([parameterization["x"+str(i)] for i in range(dim)],dtype=dtype, device=device).reshape((1,dim))
y = float(object_func(x,**function_kwargs))
y_max = max(y,y_max)
print(y,y_max)
return {"objective": (y, 0.0)}
if args.method=="alebo":
embed_dim = 6
for i in range(1,21):
alebo_strategy = ALEBOStrategy(D=object_dim, d=embed_dim, init_size=init_samples)
wct = time.time()
cput = time.process_time()
Times = []
T_process = []
y_max = -100
best_parameters, values, experiment, model = optimize(
parameters=parameters,
experiment_name="ALEBO",
objective_name="objective",
evaluation_function=object_func_ax_wrapper,
minimize=False,
total_trials=total_budget,
generation_strategy=alebo_strategy,
)
np.save(output_path + "Time_"+str(i)+".npy",np.array(Times))
np.save(output_path + "Time_process_"+str(i)+".npy",np.array(T_process))
Y = np.array([trial.objective_mean for trial in experiment.trials.values()])
X = convert_para_to_X(experiment,Y.shape[0],object_dim)
np.save(output_path + "X_"+str(i)+".npy",X)
np.save(output_path + "Y_"+str(i)+".npy",Y)
elif args.method=="hesbo":
embed_dim = 6
for i in range(1,21):
hesbo_strategy = HeSBOStrategy(D=object_dim, d=embed_dim, init_per_proj=init_samples)
wct = time.time()
cput = time.process_time()
Times = []
T_process = []
y_max = -100
best_parameters, values, experiment, model = optimize(
parameters=parameters,
experiment_name="HeSBO",
objective_name="objective",
evaluation_function=object_func_ax_wrapper,
minimize=False,
total_trials=total_budget,
generation_strategy=hesbo_strategy,
)
np.save(output_path + "Time_"+str(i)+".npy",np.array(Times))
np.save(output_path + "Time_process_"+str(i)+".npy",np.array(T_process))
Y = np.array([trial.objective_mean for trial in experiment.trials.values()])
X = convert_para_to_X(experiment,Y.shape[0],object_dim)
np.save(output_path + "X_"+str(i)+".npy",X)
np.save(output_path + "Y_"+str(i)+".npy",Y)