/
t_two_term_approximation.py
201 lines (162 loc) · 8.65 KB
/
t_two_term_approximation.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
from __future__ import division
from builtins import range
import numpy as np
from gpkit import Variable, Model
from copy import copy
import unittest
from gpkit.tests.helpers import run_tests
from robust.twoterm_approximation import TwoTermApproximation
from robust.testing.models import gp_test_model
class TestTwoTermApproximation(unittest.TestCase):
def test_equivalent_twoterm_model(self):
gpmodel = gp_test_model()
equivalent_constraints = []
for c in gpmodel.flat(constraintsets=False):
equivalent_constraints += TwoTermApproximation.two_term_equivalent_posynomial(c.as_posyslt1()[0], 0, [], True)[1]
twoterm_gpmodel = Model(gpmodel.cost, [equivalent_constraints], gpmodel.substitutions)
self.assertAlmostEqual(gpmodel.solve(verbosity=0)['cost'],twoterm_gpmodel.solve(verbosity=0)['cost'])
def test_check_if_permutation_exists(self):
for _ in range(10):
number_of_monomials = int(np.random.rand()*15) + 3
number_of_permutations = TwoTermApproximation.total_number_of_permutations(number_of_monomials)
number_of_gp_variables = int(np.random.rand()*20) + 1
m = [np.random.rand()*10 for _ in range(number_of_monomials)]
for j in range(number_of_monomials):
for i in range(number_of_gp_variables):
x = Variable('x_%s' % i)
m[j] *= x**(np.random.rand()*10 - 5)
p = sum(m)
permutation_list = list(range(0, number_of_monomials))
list_of_permutations = []
list_of_posynomials = []
counter = 0
while counter < min(100, int(np.floor(number_of_permutations/2))):
temp = copy(permutation_list)
np.random.shuffle(temp)
if TwoTermApproximation.check_if_permutation_exists(list_of_permutations, temp):
continue
else:
list_of_permutations.append(temp)
_, data_constraints = TwoTermApproximation.two_term_equivalent_posynomial(p, 1, temp, False)
data_posynomial = [constraint.as_posyslt1()[0]*Variable("z^%s_%s" % (i, 1))
for i, constraint in enumerate(data_constraints)]
list_of_posynomials.append(list(data_posynomial))
counter += 1
# counter = 0
#
# while counter < min(100, int(np.floor(number_of_permutations/2))):
# temp = copy(permutation_list)
# np.random.shuffle(temp)
#
# flag_one = TwoTermApproximation.check_if_permutation_exists(list_of_permutations, temp)
# _, data_constraints = TwoTermApproximation.two_term_equivalent_posynomial(p, 1, temp, False)
# data_posynomial = [constraint.as_posyslt1()[0]*Variable("z^%s_%s" % (i, 1))
# for i, constraint in enumerate(data_constraints)]
# flag_two = list(data_posynomial) in list_of_posynomials
#
# assert (flag_one == flag_two)
# counter += 1
def test_bad_relations(self):
for _ in range(30):
number_of_monomials = int(20*np.random.random()) + 3
number_of_gp_variables = int(np.random.rand()*10) + 1
number_of_additional_uncertain_variables = int(np.random.rand()*5) + 1
vector_to_choose_from_pos_only = [0, 0, 1, 0, 0, 0, 0, 0, 1, 0]
m = [np.random.rand()*10 for _ in range(number_of_monomials)]
p_uncertain_vars = []
relations = {}
sizes = {}
neg_pos_neutral_powers = []
for j in range(number_of_monomials):
for i in range(number_of_gp_variables):
x = Variable('x_%s' % i)
m[j] *= x**(np.random.rand()*10 - 5)
number_of_elements_in_relation = min(number_of_monomials,
int(number_of_monomials*np.random.rand()+2))
all_elements = []
for _ in range(number_of_elements_in_relation):
element = np.random.choice(list(range(0, number_of_monomials)))
while element in all_elements:
element = np.random.choice(list(range(0, number_of_monomials)))
all_elements.append(element)
element_map = {}
number_of_element_map_elements = min(number_of_monomials - 1,
int(number_of_monomials*np.random.rand()+2))
for _ in range(number_of_element_map_elements):
element_map_element = int(np.random.rand()*number_of_monomials)
while element_map_element in element_map or element_map_element == element:
element_map_element = int(np.random.rand()*number_of_monomials)
size = int(number_of_monomials*np.random.rand())+1
element_map[element_map_element] = size
try:
relations[element_map_element][element] = size
except:
relations[element_map_element] = {element: size}
if element in relations:
relations[element].update(element_map)
else:
relations[element] = element_map
relations_copy = {}
for key in list(relations.keys()):
relations_copy[key] = copy(relations[key])
sizes[key] = len(relations[key])
counter = 0
while relations_copy:
keys = list(relations_copy.keys())
for key in keys:
if not relations_copy[key]:
del relations_copy[key]
continue
u = Variable('u_%s' % counter, np.random.random(), pr=100*np.random.random())
counter += 1
p_uncertain_vars.append(u.key)
el_pow = np.random.choice([-1, 1])
m[key] *= u**(np.random.rand()*5*el_pow)
element_keys = list(relations_copy[key].keys())
for element_key in element_keys:
m[element_key] *= u**(-np.random.rand()*5*el_pow)
relations_copy[key][element_key] -= 1
if relations_copy[key][element_key] == 0:
del relations_copy[key][element_key]
relations_copy[element_key][key] -= 1
if relations_copy[element_key][key] == 0:
del relations_copy[element_key][key]
for i in range(number_of_additional_uncertain_variables):
u = Variable('u_%s' % counter, np.random.random(), pr=100*np.random.random())
counter += 1
p_uncertain_vars.append(u.key)
el_pow = np.random.choice([-1, 1])
neg_pos_neutral_powers.append([el_pow*vector_to_choose_from_pos_only[int(10*np.random.random())]
for _ in range(number_of_monomials)])
for j in range(number_of_monomials):
m[j] *= u**(np.random.rand()*5*neg_pos_neutral_powers[i][j])
p = sum(m)
monomials = p.chop()
actual_relations, actual_sizes = TwoTermApproximation.bad_relations(p)
keys = list(actual_relations.keys())
actual_relations_mons = {}
actual_sizes_mons = {}
for key in keys:
internal_map = actual_relations[key]
map_keys = list(internal_map.keys())
map_mons = {}
for map_key in map_keys:
map_mons[monomials[map_key]] = internal_map[map_key]
actual_relations_mons[monomials[key]] = map_mons
actual_sizes_mons[monomials[key]] = actual_sizes[key]
keys = list(relations.keys())
relations_mons = {}
sizes_mons = {}
for key in keys:
internal_map = relations[key]
map_keys = list(internal_map.keys())
map_mons = {}
for map_key in map_keys:
map_mons[m[map_key]] = internal_map[map_key]
relations_mons[m[key]] = map_mons
sizes_mons[m[key]] = sizes[key]
self.assertEqual(actual_relations_mons, relations_mons)
self.assertEqual(sizes_mons, actual_sizes_mons)
TESTS = [TestTwoTermApproximation]
if __name__ == "__main__":
run_tests(TESTS)