-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathvar_coord_01_29.py
163 lines (118 loc) · 5.32 KB
/
var_coord_01_29.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
#Searching over the simulation to find its sensitivity to parameters
import Comms_framework as Comms_framework
import copy
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits import mplot3d
import math
run_weights = [0,1] #uncoupled, coupled
#HYPOTHESIS 1: Getting an early developed prototype is difficult
#Figure: Best solution value with respect to time and set size on the SAE car problem
params = {}
params['update_interval'] = 20
params['max_designs'] = 1
params['compiler_iterations'] = 1
params['compiler_starting_samples'] = 1
params['compiler_interval'] = 10
params['max_iterations'] = 1000
params['meeting_length'] = 50
params['reward_scale'] = 1
params['VC_agents'] = 2
exploration_phase_fraction = 0.01 #[0.001,0.05,0.1,0.2,0.35,0.500,0.650,0.800,0.900,0.950,1]
#num_iterations = [100,200,500,750,1000]
max_designs = [1,2,3,4,5,7,9,11]
num_tries = 20
num_iterations = params['max_iterations']
#iteration_record = np.zeros((len(exploration_phase_fraction),len(num_iterations),num_tries))
obj_record2 = np.zeros((num_iterations,len(max_designs),num_tries))
best_objs = []
best_props = []
best_targets = []
best_subobjs = []
action_hist = np.zeros((num_iterations,params['VC_agents'],len(max_designs),num_tries))
for i in range(len(max_designs)):
exploration_phase_iterations = math.ceil(num_iterations*exploration_phase_fraction)
params['max_iterations'] = 10000
max_iterations = num_iterations
params['max_designs'] = max_designs[i]
params['compiler_iterations'] = 1
params['compiler_starting_samples'] = max_designs[i]*3
for k in range(num_tries):
test_framework = Comms_framework.comms_framework(params,'variable_coord')
test_framework.problem.weights[0] = run_weights[0]
test_framework.problem.weights[1] = run_weights[1]
action = np.zeros(len(test_framework.action_space.high))
action[0] = 1
print(params['max_designs'])
if params['max_designs'] == 1:
test_framework.switch_to_integration()
for j in range(max_iterations):
if j == exploration_phase_iterations:
test_framework.switch_to_integration()
old_obj = test_framework.best_solution_value
test_framework.step(action)
step, bsv,im,aa,state_dict = test_framework.render()
good_actions = state_dict["good_actions"]
action_hist[j,:,i,k] = good_actions
new_obj = test_framework.best_solution_value
obj_record2[j,i,k] = new_obj
best_props.append(test_framework.best_props)
best_targets.append(test_framework.best_targets)
best_objs.append(new_obj)
best_subobjs.append(test_framework.problem.global_objective(test_framework.best_props,test_framework.best_targets))
r_2 = obj_record2
np.save('model_00_01w_2ag',obj_record2)
params = {}
params['update_interval'] = 20
params['max_designs'] = 1
params['compiler_iterations'] = 1
params['compiler_starting_samples'] = 1
params['compiler_interval'] = 10
params['max_iterations'] = 1000
params['meeting_length'] = 50
params['reward_scale'] = 1
params['VC_agents'] = 9
exploration_phase_fraction = 0.01 #[0.001,0.05,0.1,0.2,0.35,0.500,0.650,0.800,0.900,0.950,1]
#num_iterations = [100,200,500,750,1000]
max_designs = [1,2,3,4,5,7,9,11]
num_tries = 20
num_iterations = params['max_iterations']
#iteration_record = np.zeros((len(exploration_phase_fraction),len(num_iterations),num_tries))
obj_record2 = np.zeros((num_iterations,len(max_designs),num_tries))
best_objs = []
best_props = []
best_targets = []
best_subobjs = []
action_hist = np.zeros((num_iterations,params['VC_agents'],len(max_designs),num_tries))
for i in range(len(max_designs)):
exploration_phase_iterations = math.ceil(num_iterations*exploration_phase_fraction)
params['max_iterations'] = 10000
max_iterations = num_iterations
params['max_designs'] = max_designs[i]
params['compiler_iterations'] = 1
params['compiler_starting_samples'] = max_designs[i]*3
for k in range(num_tries):
test_framework = Comms_framework.comms_framework(params,'variable_coord')
test_framework.problem.weights[0] = run_weights[0]
test_framework.problem.weights[1] = run_weights[1]
action = np.zeros(len(test_framework.action_space.high))
action[0] = 1
print(params['max_designs'])
if params['max_designs'] == 1:
test_framework.switch_to_integration()
for j in range(max_iterations):
if j == exploration_phase_iterations:
test_framework.switch_to_integration()
old_obj = test_framework.best_solution_value
test_framework.step(action)
step, bsv,im,aa,state_dict = test_framework.render()
good_actions = state_dict["good_actions"]
action_hist[j,:,i,k] = good_actions
new_obj = test_framework.best_solution_value
obj_record2[j,i,k] = new_obj
best_props.append(test_framework.best_props)
best_targets.append(test_framework.best_targets)
best_objs.append(new_obj)
best_subobjs.append(test_framework.problem.global_objective(test_framework.best_props,test_framework.best_targets))
r_2 = obj_record2
np.save('model_00_01w_9ag',obj_record2)